You are on page 1of 576

TURBINE

STEAM PATH
MAINTENANCE AND REPAIR
Volume 1

William P. Sanders, P. Eng.


Disclaimer: The recommendations, advice, descriptions, and the
methods in this book are presented solely for educational purposes.
The author and publisher assume no liability whatsoever for any loss
or damage that results from the use of any of the material in this book.
Use of the material in this book is solely at the risk of the user.

Sanders, William P. Turbine Steam Path Maintenance and Repair Volume


One / William P. Sanders, P.E.
p. cm. q.cm
Includes index

ISBN 0-87814-787-X

ISBN13 978-0-87814-787-8

Copyright 2001 by PennWell Corporation 1421 South Sheridan Road


Tulsa, OK 74112

800-752-9764 sales@pennwell.com www.pennwell-store.com


www.pennwell.com

Cover and book design by Robin Brumley

All rights reserved. No part of this book may be reproduced, stored in a


retrieval system, or transcribed in any form or by any means, electronic
or mechanical including photocopying or recording, without the prior
written permission of the publisher.

Library of Congress Cataloging-in-Publication Data

Printed in the United States of America

2 3 4 5 6 12 11 10 09 08
Turbine Steam Path Maintenance and RepairVolume One

PREFACE

The Turbine Steam Path, Damage,


Deterioration, and Corrective Options
This book has been prepared for those technical people respon-
sible for the operation and maintenance of steam turbines.

Steam turbines represent a complex technology for units com-


monly designed to operate hundreds of thousands of hours while
being subjected to a severe environment and a variety of operating
phenomena capable of degrading their condition. These units are
required to continually operate in a reliable, safe, and cost effective
manner. Under such circumstances, these units cannot maintain
their original design-specified level of performance indefinitely. All
units will deteriorate with age. Owners anticipate this, and designers
will normally leave an adequate margin, knowing that some level of
such deterioration is tolerable.

The technology of steam turbineswhile maturecontinues to


evolve. More accurate and time-responsive diagnostic tools and
techniques are becoming available to assist in predicting when a unit
has deteriorated to the extent that corrective action is required.
Similarly, tools are available to assist the operator in analyzing prob-
lems and determining the effective corrective action best suited to
the condition causing deterioration. The improved understanding of
unit condition and rates of deterioration now achieved, together with
advances in materials, should allow units to be maintained in a man-
ner that will help minimize maintenance concerns and costs.

It is the premise of this book that units as supplied will fulfill


two basic requirements:

xii
Preface

It is assumed the unit as designed represents an optimum


selection of component sizing and arrangement

It is assumed the unit as delivered meets design specifica-


tion within the range of tolerances provided by the design
engineer, i.e., unit components have been manufactured,
assembled, tested, and installed in such a way that they are
in compliance with the original design specification

The implication of this second assumption is that if nonconform-


ing situations or conditions arose during the total manufacturing
process (and exist within the unit), they have been evaluated by a
competent design authority in the engineering organization of the
manufacturing company and have been assessed as not having an
adverse impact on the potential performance of the unit.

In terms of turbine unit components, design optimum is a dif-


ficult term to define. The entire design process is one of compromise
by the designer who wants a unit to be both efficient and reliable.
These requirements often represent competing demands, forcing the
designer to select from among various elements, possibly electing to
downgrade one aspect of these requirements to meet the demands of
the other. This is done consciously and with detailed evaluation to
provide a balanced selection.

Units delivered by a manufacturer represent the supply of ele-


ments that conform to the design principles established by his or her
design function, and conform with the best technology available to
that supplier at the time the design specification was prepared.
However, the operator must recognize that the labor and material
costs involved in building a steam turbine are high, and turbine sup-
pliers must be able to produce units at competitive levels sufficient
to allow them to achieve a profit margin enabling them to sustain
business as well as finance further development.

xiii
Turbine Steam Path Maintenance and RepairVolume One

Many power systems are currently experiencing significant


changes in how they operate. Pressures from deregulation, environ-
mental concerns and legislation, and an aging fleet of power gener-
ating equipment is shifting emphasis from the installation of new
capacity to the maintenance and care of the old. There is a continu-
ing increase in demand for electric power but new capacity installa-
tion is not keeping up with it. Operators of turbine generators are
therefore required to meet this demand with their existing fleets
aging units requiring greater care to reduce the possibility of forced
outages. The prospect of units experiencing extended outages as
damage is found at planned outages.

Historically, as units have aged they have tended to be used less


frequently. They are initially placed on spinning reserve and ulti-
mately placed in reserve, mothballed, or retiredtheir capacity
replaced with newer, more efficient units. An advantage of this dwin-
dling reserve is that older units have continued to operate at high
load factors and therefore become less susceptible to the rigors of
start-up, shut down, and the associated thermal transients.
Unfortunately, there have also been fewer opportunities for plant
maintenance to proceed with the maintenance outages required to
maintain unit operational health.

Maintenance problems associated with keeping aging units avail-


able are only going to increase. Operators who are expected to pro-
vide power on demand are going to experience even greater future
challenges of damage and deterioration. They will be expected to
identify not only the damage, but also the causative effects, and then
find immediate solutions that will not jeopardize system security.

This book examines the damage deterioration and failure mecha-


nisms occurring with unfortunate consequenceson some units,
with monotonous regularitywithin the turbine steam path. These
various forms of degradation can be the result of a number of factors
related to conditions often beyond the control of operating and main-
tenance personnel. However, even if the steam turbine is operated

xiv
Preface

precisely as intended by design, and suffers no external degrading


effects for its entire operating life, the steam environment is one that
can cause components to suffer various forms of distress. Under nor-
mal circumstances, the design process selects and defines individual
components suitable for the design operating life of the unit (normal-
ly about 200,000 hours). At a mean load factor of about 75%, this
represents a 30-year operating life.

A number of unavoidable influences affect the operating life of


the various components comprising the turbine. These include the
steam environment itself, the stresses induced in the components by
rotation, and stresses induced in various portions of the unit by
expansion of the steam through the blade passages. There are also
the effects of the high- pressure steam, causing high-pressure drops
across some components that must be contained by the casings.

External factors that can affect the reliability of components of


the steam path and act to lower the expected operating life include
the possible formation of corrosive elements at various locations
within the steam cycle, or impurities gaining access from in-leakage
at sub-atmospheric pressures. There can be unit trips caused by a
number of circumstances, from system trip electrical faults to light-
ning strikes on power lines. Many of these factors, while possibly
occurring in a 30-year operating life, cannot be anticipated in terms
of when, where, how many, or how severe their effects might be.

The damage and deterioration that occurs within the steam path
can be of several forms. It can result in a gradual material lossthe
growth of a crackor an immediate failure causing a forced outage.
Gradual deterioration can (depending upon type and location) be
monitored and replacement parts made available, or corrective
action taken to rectify the situation before it extends to an unaccept-
able degree. Immediate failure is most often the consequence of
either mechanical rupture or the presence in the steam path of some
foreign object, either generated within or having gained access from
some external source (including drop-ins).

xv
Turbine Steam Path Maintenance and RepairVolume One

In writing this book, I have tried to present information that plant


personnel will be able to use to make value judgments on the type
and severity of any damage, suggest possible causes, and then con-
sider the most appropriate corrective actions that are available. To
aid in the recognition and classifying of operational damage and
deterioration, photographs are used to illustrate unacceptable or
suspect conditions.

Many of the damaging phenomena considered in these chapters


do not occur in isolation. It is possible that several can and will
occur simultaneously, demonstrating that components are subjected
to more than one degrading influence. A condition may initiate due
to one damaging mechanism introducing a condition of weakness,
which then allows another mechanism to become predominant and
drive a component to failure. This situation often occurs even though
the driving mechanism would not have been capable of causing fail-
ure had not the weakness been introduced by the first, or initiating
mechanism.

Before considering degradation and failure in any detail, it is


important to define what constitutes failure and/or deterioration. An
important consideration in any case of evaluation and condition
assessment of a turbine is establishing what constitutes failure. The
definition I find most acceptable is this: A condition exists within the
unit that while it would not prevent the unit from returning to serv-
ice and continuing to develop power, it could force it from service
before the next planned outage. Various other definitions exist, and
the definition of failure used in any situationand therefore the
responsibility for correctioncan be controversial. This controversy
is to some extent aggravated by possibilities; e.g., a crack that has
been determined to exist may be predicted by the methods of frac-
ture mechanics to be growing at a rate that would not cause com-
plete rupture, forcing the unit from service before the next planned
outage.

xvi
Preface

As reserve power margins diminish, steam turbinesthat cur-


rently have operating periods between major maintenance outages
of three to eight yearscould be forced to operate longer than
intended when they were originally returned to service. Under these
circumstances, it is difficult when making a prediction of a units
future operation, to be certain there will not be some major change
in its operating parameters. Parameters that can influence an accept-
able definition of failure in any situation include the exact operating
period, the unit load pattern, and the steam conditions the unit will
experience over a number of years.

A simple and conservative solution to this definition of failure


would be to change any suspect component showing any crack or
unacceptable damage-or-deformation indication. This may appear to
be an expensive option, but is considerably less expensive than a
forced outage requiring weeks or months to open, repair, await
replacement parts, replace those parts, close the unit, and return it to
service.

Defining efficiency deterioration is somewhat easier. It is even


possible to quantify such deterioration in terms of reducing steam
path efficiency and unit output. What is not possible to determine is
the extent of any mechanical deterioration that may occur and cause
efficiency deterioration. This is an unknown situation not recognized
until complete mechanical rupture occurs. There is normally no
manner to predict such an occurrencedamage could be in the
incubation phaseeven when an examination of the steam path is
made at maintenance outages.

During operation, certain situations and phenomena are known


to occur that have the potential to initiate damage or to cause dete-
rioration in performance. These damaging and deteriorating phe-
nomena can be of a continuous or intermittent nature, produced as
a consequence of transient operating or steam conditions. Such phe-
nomena can also be the result of sudden mechanical failures of com-
ponents that cause more extensive consequential damage. The most

xvii
Turbine Steam Path Maintenance and RepairVolume One

commonly occurring of these degrading effects are related to the for-


mation of moisture in the steam path or solid foreign particles, pos-
sibly from the boiler or scale generated within the superheater and
reheater tubes. Other sources include chemical contaminants that
are introduced, or gain access to the steam path on which they are
deposited, and possibly act as corrosive elements. The other princi-
pal degrading condition is the operational phenomena occurring
during the operating life of the unit.

The first two chapters of this book provide general information.


The first outlines what is considered necessary to define and consti-
tute a maintenance strategy that represents managements commit-
ment to maintaining a healthy system. This chapter also outlines
means of monitoring conditions indicative of damage. The second
chapter deals with the spatial arrangement within the steam path and
the factors that affect it. This is important because the performance
(efficiency and reliability) of a turbine is influenced considerably by
the alignment of the unit and the resulting axial and radial clear-
ances and laps that are achieved in the hot operating condition.

Chapters 3, 4, 5, and 6 discuss the various phenomena known to


affect both the efficiency and structural integrity of the components.
In the second volume, chapters 7, 8, and 9 consider repair and refur-
bishment options currently available. Fortunately, there are ever-pres-
ent advances in these technologies, and as experience is gained,
newer and improved methods develop to be applied to older units so
they can continue to operate with high levels of availabilityoften
with improved efficiency. Chapter 10 considers seal systems and
gland rings, and provides means of estimating the financial penalties
associated with excessive leakage. Seals are one area where opera-
tors and maintenance personnel can influence the cost of power gen-
eration, and help reduce the cost of power to their customers.

The final two chapters, 11 and 12, relate to quality and the inspec-
tion of elements being manufactured to replace damaged compo-
nents. This is an area where many engineers feel the cost of undertak-

xviii
Preface

ing such inspections is difficult to justify. However, what happens


when componentsmanufactured when they are required in an
emergency to return a unit to servicehave any form of fault and
force the unit from service prematurely? In such a case, the cost of
inspectionensuring that a suppliers quality program is prepared and
operating properlyis well justified. It is often said, There isnt time
and money to do it right, but there is always time and money to cor-
rect it. This statement is well applied to the manufacture or repair of
components in an emergency, because the cost of a second outage is
just as high as the first, and far more embarrassing.

Because the steam turbine is a thermal machine designed to con-


vert thermal energy to rotation kinetic energy, I have included an
appendix that provides the basic thermal relationships required to
understand the turbine and its operation.

Situation evaluation
The more susceptible areas in any turbine unit are a function of
many complex factorsindividual stress levels, stress concentration,
mode of operation, and the operating environment. Individual com-
ponents are also greatly influenced by the expertise with which the
parts were designed, manufactured, and assembled, and the oper-
ating transients to which they have been subjected. The diversity of
the factors that can contribute to damage precludes any generaliza-
tion of cause or value. Steam path components are subjected to high
stress, both direct and alternating. Many parts operate at high tem-
peratures and are of complex forms interacting with one another in
unpredictable ways. These factors, when combined with load and
temperature transients that occur during operation, combine to make
the steam path highly sensitive and a major source of concern to the
designer and operator.

While some concerns are common to most operators, the type of


deterioration or damage to which any component or area is subjected

xix
Turbine Steam Path Maintenance and RepairVolume One

normally varies from unit to unit. This accounts for the variety of con-
cerns expressed by maintenance staff, and the different dispositions of
the various nonconforming conditions that will be developed in any
situation.

In many instances when corrective action is required, there is no


optimum solution that can be followed without deviation. Operation
and load demands will often negate the optimum. At other times,
costs, special tools, skills, and the availability of replacement parts
could require some form of compromise. These compromise solu-
tions may have to be adopted from necessity, but the final disposi-
tion should provide the best balance between cost, risk, and the
immediacy of returning the unit to service.

The logical approach to maintenance and repair dispositions is:

Consider the available alternatives in terms of the original


design requirements of the affected components

Evaluate possible solutions in terms of departure from the


design specified requirements

Many repair or accept-as-is dispositions will have only a lim-


ited effect on unit performance, and can be readily accepted. Other
repairs can be proposed and accepted, representing a compromised
condition. Such options should only be accepted on the basis that
the unit will be operated with this compromised solution for as short
a period as possible, and that the selected option does not represent
a significant level of risk in the short term. If this is possible, plans
should be put into effect immediately to develop an acceptable solu-
tion that can be undertaken within a reasonable time.

The maintenance options


The satisfactory performance of a steam turbine is influenced con-
siderably by the manner and expertise with which it is maintained,

xx
Preface

and the load patterns it follows. While the plant operating engineer
can control, to a large degree, the maintenance of the units for which
he is responsible, he is unfortunately unable to exercise little influence
on operating patterns. This is a responsibility of dispatchers who have
a mandate to serve the demands of their clients rather than the turbine
generators of their system.

For maintenance to be cost-effective, it must be planned. When


signs of distress, excessive wear, misalignment, or component dete-
rioration are detected, the need for corrective action must be con-
sidered. These corrective actions should help ensure the situation
does not deteriorate further, to the extent the unit is placed on a
forced outage status, severely load limited, or suffers an unaccept-
ably high degree of deterioration in efficiency.

There are general maintenance requirements for any unit.


Guidance for these is provided by the designer and should be fol-
lowed for all routine matters. The designer will also provide recom-
mendations for the operating time between opening sections of the
unit for periodic maintenance and examination. During these main-
tenance outages, any findings that could affect unit performance
must be reviewed in relation to their possible long-term effects.

Maintenance actions
Opening a unit for maintenance provides the opportunity to
make repairs or to install replacement parts when the necessary skills
and special purpose tools are available. Such an opening also allows
replacement parts to be ordered, which can be placed in the unit at
the current or later outage, depending upon the delivery and
required period of the outage. Replacement is made when an evalu-
ation of any found operational nonconformance is judged to be
placing the unit at risk if returned to service without correction. A
detailed evaluation of each nonconformance should be made and it
should indicate if, and what actions are required.

xxi
Turbine Steam Path Maintenance and RepairVolume One

The principal purpose of a steam turbine maintenance inspection


is to detect potential problems at an early stage. If this is not done,
relatively minor situations could progress to the extent a forced out-
age or excessive loss in unit output and efficiency could occur.
During such a maintenance inspection outage, parts can be exam-
ined visually for indications of failure, wear, or distortion. Also, non-
destructive tests can be applied to critical components to determine
if their ability to continue to perform satisfactorily has deteriorated
and if so, what remedial action should be taken, or planned.

A nonconformance in any part of the steam turbine unit is con-


sidered to have occurred when there are signs of mechanical failure,
excessive wear, or any form of deterioration that has the potential to
adversely affect the performance of the unit. Such nonconformances
must be reviewed for its short- and long-term effects.

As soon as unit inspection indicates a nonconforming condition


has been found, it must be evaluated. The logic process of evaluation
for both availability and efficiency is considered in chapter 1. This
chapter outlines avenues the maintenance engineer should explore
in deciding what corrective action needs to be taken. There are four
decisions that can be reached. In some circumstances the decision
is relatively simple, and is in fact obvious. In other situations, a deci-
sion is made based on the probability of failure, the possible cost of
repair, and ultimately, the reparation of consequential damages that
are the result of not taking corrective action. These four options can
be considered:

scrap and replace

repair

rework

accept-as-is

xxii
Preface

Of these decisions, possibly the most difficult and potentially


most controversial is the latteraccept as isa disposition that
allows a component to return to service with no effort made to cor-
rect the nonconforming condition. There are two reasons for reach-
ing and deciding upon this course of action:

There is little need to make any corrections. To make them


will add no or marginal improvement to unit performance
and the condition will not place the unit at risk

The cost of replacing, repairing, or reworking cannot be jus-


tified. This is often a judgment call on the part of the engi-
neer and can only be made if he or she is aware of any risks
involved

Such a decision should not be made as a desperation measure.


The risks, if any, should be fully evaluated. The options and the prob-
ability of failurefrom an extended outage to operationmust be
fully considered.

Therefore, the evaluation process can be a complex one.


Occasionally, the solution is self-evidentsuch as when partial fail-
ure has occurred, or when excessive damage exists. The most diffi-
cult decisions are those related to suspected damage or deteriora-
tion, and those for which it is difficult to determine the cause. In
these instances of uncertainty, mature judgment is required, togeth-
er with knowledge of the operating and maintenance history of the
unit. This knowledge should help in the evaluation. The information
in this book can also provide confidence in the selection of the final
disposition.

The availability of replacement parts, special skills, and tools will


often influence which decision is reached. Care must be exercised to
ensure that availability or non-availability of replacement parts does
not force the owner/operator into a decision ultimately causing more

xxiii
Turbine Steam Path Maintenance and RepairVolume One

expense and increasing the overall risk level to an unacceptable


degree.

Often, alternatives to these potential solutions are available.


Some may degrade a units rating or impose other restrictions in
terms of maximum output, or the time for which a unit can be oper-
ated. The compromise correction is ultimately more acceptable over
the short-term, while the owner/operator arranges for a more palat-
able long-term solution.

William P. Sanders
Richmond Hill, Ontario, Canada
August, 1999

xxiv
TABLE OF CONTENTS

List of Acronyms . . . . . . . . . . . . . . . . . . . . . ix

Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . x

Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . xii

Acknowledgements . . . . . . . . . . . . . . . . . . xxv

Chapter 1Considerations of a Turbine Steam


Path Maintenance Strategy
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Considerations Relating to a Maintenance Strategy . . . . . . . 2
The Turbine Outage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
Establishing the Need for Unit Shutdown . . . . . . . . . . . . . . 7
Outage Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
Interval Between Maintenance Outages . . . . . . . . . . . . . . 13
The Inspection/Maintenance Outage . . . . . . . . . . . . . . . . 17
The Available Corrective Options . . . . . . . . . . . . . . . . . . . 19
Distinction Between Causes and Mechanisms of Failure . . 25
Component Susceptibility for Deterioration . . . . . . . . . . . . 52
Instantaneous Damage or Failure . . . . . . . . . . . . . . . . . . 57
Factors Contributing to Gradual Deterioration . . . . . . . . . . 59
Monitoring Damage and Deterioration . . . . . . . . . . . . . . . 66
Replacement Parts Strategy and Supply . . . . . . . . . . . . . . . 82
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

v
Turbine Steam Path Maintenance and RepairVolume One

Chapter 2Steam Path Component Alignment


and Stage Spatial Requirements
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
Predictable Factors Affecting Design Clearance . . . . . . . . . 93
Rotor Vertical Deflection . . . . . . . . . . . . . . . . . . . . . . . . . 95
Differential Expansion . . . . . . . . . . . . . . . . . . . . . . . . . . 108
Radial Expansion of the Steam Path Parts . . . . . . . . . . . . 145
Diaphragm Deflection at Pressure and Temperature . . . . 159
Unit End-to-End Lateral Alignment . . . . . . . . . . . . . . . 166
Methods of Field Alignment . . . . . . . . . . . . . . . . . . . . . . 167
Unpredictable Factors Affecting Design Clearance . . . . . . 175
Steam Path Area Requirements . . . . . . . . . . . . . . . . . . . . 188
The Stage Operating Definition . . . . . . . . . . . . . . . . . . . 190
Steam Path Component Arrangement (Axial/Radial Direction) . 192
Blade Vane and Cascade Geometry . . . . . . . . . . . . . . . . 229
The Effect of Vane Placement Errors . . . . . . . . . . . . . . . . 248
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282

Chapter 3Steam Path Damage


Induced by Water
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
Water Condensation in Expanding Steam . . . . . . . . . . . . 285
Radial Distribution of Moisture . . . . . . . . . . . . . . . . . . . 289
Moisture Deposition . . . . . . . . . . . . . . . . . . . . . . . . . . . 292
Measuring Moisture Distribution and Content . . . . . . . . . 301
Water Removal from the Steam Path . . . . . . . . . . . . . . . . 304
Moisture-induced Damage . . . . . . . . . . . . . . . . . . . . . . . 322
Moisture-impact Erosion . . . . . . . . . . . . . . . . . . . . . . . . 323
Blade-trailing Edge Erosion . . . . . . . . . . . . . . . . . . . . . . . 352
Wire-drawing Erosion . . . . . . . . . . . . . . . . . . . . . . . . . . 370
Water-washing Erosion . . . . . . . . . . . . . . . . . . . . . . . . . 379
Water Ingestion into the Steam Path . . . . . . . . . . . . . . . . 389
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395

vi
Table of Contents

Chapter 4Operational Events


Giving Rise to Steam Path Damage
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
Foreign Object Impact Damage . . . . . . . . . . . . . . . . . . . 398
Sources of the Impacting Objects . . . . . . . . . . . . . . . . . . 401
Impact Damage Classification . . . . . . . . . . . . . . . . . . . . 409
Solid-particle Erosion (Abrasion) . . . . . . . . . . . . . . . . . . . 426
Scale Formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 428
The Erosion Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . 431
Material Loss Patterns Due to SPE . . . . . . . . . . . . . . . . . . 439
Protective Measures Against Erosion . . . . . . . . . . . . . . . . 459
SPE Influence on Stage Performance . . . . . . . . . . . . . . . . 460
Steam Path Component Rubbing . . . . . . . . . . . . . . . . . . 462
Fretting Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 477
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486

Chapter 5Steam Path Damage


and Deterioration from Material
Property Degradation
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
Considerations of Material Structure . . . . . . . . . . . . . . . . 489
High-temperature Creep . . . . . . . . . . . . . . . . . . . . . . . . 490
Creep Deformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 493
The Creep Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . 497
Creep Rate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 504
Creep in Steam Path Components . . . . . . . . . . . . . . . . . . 507
High-cycle Fatigue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 520
The High-cycle Phenomena . . . . . . . . . . . . . . . . . . . . . . 522
Rotating Blade Vibratory Stresses . . . . . . . . . . . . . . . . . . 525
Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
Fatigue Stresses and their Representation . . . . . . . . . . . . 569
Crack Growth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576
HCF Failure Surface Appearance . . . . . . . . . . . . . . . . . . 579

vii
Turbine Steam Path Maintenance and RepairVolume One

Creep Fatigue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 581


Temper Embrittlement . . . . . . . . . . . . . . . . . . . . . . . . . . 582
Low-cycle Thermal Fatigue . . . . . . . . . . . . . . . . . . . . . . . 583
Thermal Transients . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 586
Determination of Thermal Stresses . . . . . . . . . . . . . . . . . 596
Components Operating at High Temperature . . . . . . . . . . 600
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 604

Chapter 6
Steam Path Damage and Deterioration from
the Deposition of Contaminants
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607
Source of Steam Path Impurities . . . . . . . . . . . . . . . . . . . 610
The Composition of Deposits . . . . . . . . . . . . . . . . . . . . . 619
The Removal of Chemical Deposits from the Steam Path . . 637
Steam Path Cleaning Methods . . . . . . . . . . . . . . . . . . . . 639
Deposition Patterns . . . . . . . . . . . . . . . . . . . . . . . . . . . . 644
Steam Path Efficiency Deterioration . . . . . . . . . . . . . . . . 659
Steam Path Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . 671
Forms of the Corrosion Process . . . . . . . . . . . . . . . . . . . 677
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 704

viii
Chapter

1
Considerations of a
Turbine Steam Path
Maintenance Strategy

INTRODUCTION
Although most parts of the steam turbine are capable of suffering
mechanical damageand do sustain itsome areas or components
suffer greater levels of deterioration than others. Why some areas in
any unit are more susceptible than others is a function of many com-
plex factorsindividual stress levels, stress concentration, the mode
of operation, operating environment, and the manner in which the
unit is maintained. Other critical factors involve the operating tran-
sients to which components are subjected. This diversity of factors
that can influence the potential for damage precludes any ability to
state a generalization of causes or of value. Despite this fact, the area
of the unit having a considerable potential to affect performance
and of raising the concern of the operating and maintenance engi-
neeris the steam path.

1
Turbine Steam Path Maintenance and RepairVolume One

Components of the steam path are subjected to high stresses,


direct and alternating. Many parts operate at high temperatures.
Often they are complex. These elements will also interact with oth-
ers in unpredictable ways. These factors, when combined with load
and steam temperature transients that commonly occur during oper-
ation, make the steam path highly sensitive and a prime candidate
for deterioration. If populations of steam turbines were to be
reviewed, there is a high probability that failures or problems within
the steam path would be a significant cause of deteriorating per-
formance for many of them. The rotating blades would pose a major
cause within the steam path itself.

This chapter discusses strategies available to owners when they


determine their units have suffered some form of structural deterio-
ration within the steam path (to the extent corrective action is con-
sidered necessary, or at least considered prudent to evaluate the con-
dition before the unit is returned to service).

When it is determined that a nonconforming situation (damage)


exists within the steam path, the user must consider available
options. In these circumstances, it is necessary to evaluate the situa-
tion and decide which action will properly utilize the technical
capabilities and skills available to the operator. Decisions relating to
the condition and the evaluation include considerations of both cost
and time, and the potential costs of electing not to take corrective
action during the current outage.

CONSIDERATIONS RELATING TO
A MAINTENANCE STRATEGY
In order to operate turbine units in the most cost-effective man-
ner, each operating organization must define a specific policy gov-

2
Consideration of a Turbine Steam Path Maintenance Strategy

erning the maintenance of T-G units. It is normally inefficient and


very costly to operate units until they are forced from service and
maintenance undertaken on an emergency basis. An effective pol-
icy must recognize that some spare capacity must exist within the
system, and that power must be able to be purchased or exchanged
between neighboring and interconnected systems so units can be
shut down for periods sufficient to allow them to be opened, exam-
ined, and where necessary, corrective action taken.

An effective maintenance strategy for any population of steam


turbines owned and operated by a single authority carries certain
requirements:

The strategy and the corrective actions that an evaluation for


any condition selects must be cost effective

The selected corrective actions (if actions are required) must


be able to be performed within a reasonable time frame

The selected corrective actions must not create a condition


that will in any way compromise the availability of the unit

The selected corrective actions must not modify the physical


characteristics of the steam path components to the extent
they cause a deterioration of unit efficiency beyond what is
determined to be acceptable

The selected action must be able to be performed by main-


tenance personnel available from within the plant or spe-
cialized outside workers. Any required, specialized tools
must be available as needed

Considerations of three separate outages influence (or are influ-


enced by) the work scope and corrective actions at each maintenance
outage, as shown in Figure 1.2.1, i.e., the preceding outage, the one
being entered, and the next one planned (presumably in four or more

3
Turbine Steam Path Maintenance and RepairVolume One

years time) have to be considered. The major objectives of the planned


outage and subsequent inspection can therefore be considered:

To undertake normal maintenance (cleaning and replace-


ment of consumable parts of the unit, such as seal strips,
which upgrade or restore the efficiency level)

To determine the units condition and the possible need for


corrective action at some future outage

To take corrective action on conditions noted at the previous


outage. These previous conditions may have been judged at
that time to be acceptable and the unit returned to service
with a recommendation for deferred action

To record measurements and other conditions being moni-


tored to determine the rate of deterioration of any condition
that could at some future time require corrective action

To take emergency corrective action on conditions found


since the last inspection and discovered upon opening the
unit. These are conditions that could prevent the unit being
returned to service with a probability of performing at an
acceptable level

If the condition(s) discovered could prevent the unit from being


returned to service, there are three possible courses of action avail-
able. Each must be separately evaluated:

To replace the damaged or deteriorated parts. To do this,


replacement parts must either be available in inventory or on
just in time delivery. The owner must evaluate the cost of
extending the outage for hard-to-get parts

To refurbish components as needed. This action is only pos-


sible to the extent that components are capable of refurbish-
ment. There are circumstances that would allow permanent

4
Consideration of a Turbine Steam Path Maintenance Strategy

repairs and others that would allow refurbishment sufficient


for the unit to be returned to service until the next outage,
when replacement parts could be installed

To remove affected parts. This can sometimes allow the unit


to operate at a de-rated condition. However, even to do this
often requires the installation of components such as pres-
sure reducing plates (to correct pressure distribution through-
out the unit to meet more closely the design conditions),
which will allow the unit to operate without further unac-
ceptable deterioration

The action the owner selects depends upon economic consider-


ations of the total situation. This requires a careful evaluation of the
options available under the actions items listed above. This evalua-
tion should determine the most economical solution, consistent with
returning the unit to a satisfactory mechanical condition.

Replacement Special Specialized


Parts. Tools. Skills.

Input data from MAINTENANCE Data to Next


Previous Outage. OUTAGE. Planned Outage.

Order Replacement
Dimensional
Parts.
Records.
Comparison of Condition
and any Deterioration.

Fig. 1.2.1The results of three openings impact on each outage, the previous in terms
of the preparation for the present, the present for corrective actions identified for the
future, and also to develop plans and work scopes for repair or refurbishment at the
next.

5
Turbine Steam Path Maintenance and RepairVolume One

THE TURBINE OUTAGE


Maintenance or planned outage
This outage is scheduled. It is part of the established plan for
maintaining the turbines in an acceptable condition, and may
require special skills and tools be made available to undertake the
required work. These outages may also be an opportunity to replace
parts that were damaged or badly worn and noted as damaged or
deteriorated at a previous outage. Replacement parts would have
been specially ordered for this outage, and its occurrence could have
been time dependent upon their delivery.

Forced outage
This is an unplanned outage that occurs as a consequence of
some unexpected failure or damage occurring within the turbine
unit. The damage may result in a condition indicating an unaccept-
able situation such as high vibration, noise, or even a unit trip. If
such an incident occurs, operators must decide whether to continue
in the existing mode or shut down the unit and investigate. In the
event of a unit trip, the operators may attempt to re-synchronize the
unit and continue to operate.

To shut down the unit and investigate can be an expensive


option in terms of material and labor costs associated with opening
the unit, as well as the lost generating capacity for the time the unit
is not producing power. To not shut down the unit can be even more
expensive, however, if such action neglects a condition that is in the
early stages of development, and which has caused the early system
interruption in the first place. An uncorrected condition could
engender more serious damage to other components within the unit,
and could ultimately cause extensive damage or force major com-
ponents to be scrapped. Safety has to be considered in any decision

6
Consideration of a Turbine Steam Path Maintenance Strategy

not to shut down a unit. A condition may exist that indicates a major
and possibly catastrophic failure may occur.

Extended outage
An extended outage is the result of finding a situation during a
planned outage that causes the planned time frame for the unit
inspection, cleaning, and refurbishment to be extended. Normally
this occurs when damage or deterioration is found requiring the use
of nonstandard tools or skills for evaluation, or the necessary
replacement parts being unavailable.

ESTABLISHING THE NEED


FOR UNIT SHUTDOWN
Outside of the normal period during which a unit is returned to
service and expected to operate at an adequate level of perform-
ance, there are certain indications from a unit denoting a need to
consider the situation, investigate it, and possibly shut down for an
unscheduled outage.

The signals indicating damage or deterioration


Station operating procedures include the need to monitor condi-
tions of certain operating parameters and, when they vary outside
expected norms, can indicate the need for emergency corrective
action. The more common of these are shown in Figure 1.4.1, and
include:

7
Turbine Steam Path Maintenance and RepairVolume One

Vibration levels. An increase in the vibration levels of the unit


rotors is an indication problems exist. There are various possible
causes for this amplitude increase. Among the more common are:

A change in the unit alignment. This loss of alignment may


be caused by various phenomena, and the possible causes
will need to be investigated separately. A vertical or horizon-
tal shift of the bearings position can cause such a situation. If
this has occurred it will be necessary to realign the shaft. The
extent to which the unit must be opened and rotors and other
components dismantled will depend upon its configuration
and the access that can be made to the bearings

The mechanical rupture of a component that has disrupted


the dynamic balance of the unit. Such mechanical failure
will not necessarily cause an imbalance. Shorter blades or
whole portions of a coverband may detach without signifi-
cantly affecting the balance and therefore the vibration level

Heavy rubs that have caused a shaft bend. If there has been a
heavy rub caused by some transient or other condition, it is
quite possible the shaft will bend to the extent balance is upset.
If this has occurred it becomes necessary to open the unit,
remove the affected rotor, and undertake major restoration

Starvation of oil to the bearings (axial or thrust). If this situa-


tion occurs it could cause significant, and possibly irrepara-
ble damage. This situation should be rectified as quickly as
possible

Noise. One of the most significant and useful indications of dam-


age or pending failure is noise from the unit. Because of the high
rotations speeds, even the lightest rubs can be indicative of pending
problems. Often these noises appear and then disappear, so it should
be noted if the noises are associated with transient conditions, e.g.,
changes in steam conditions or load. In any event:

8
Consideration of a Turbine Steam Path Maintenance Strategy

If possible, determine if the noise is sensitive to transient con-


ditions, speed, and temperature

Identify the location of the rub as accurately as possible

If the noise is continuous, try to determine if it is at a constant


level and frequency. This is subjective judgment, but careful
observation can assist

When the unit is opened, examine the region where the noise
appeared to be centered and look for rubbing damage in the form of
grooves, lifted coverbands, and changes in material hardness. There
is often a change in hardness associated with rubs that could cause
later failurethe material structure on the steel will have changed,
making it more brittle.

Fig. 1.4.1Indications of a distress or a damage


condition.

A reduction in unit output for a given steam input. If there has


been a step change in the output for a given steam input, this can be
indicative of mechanical damage in the blade systemeither a
blockage early in the steam path, or some broken elements that are
reducing the efficiency of energy conversion. Often the region where

9
Turbine Steam Path Maintenance and RepairVolume One

this damage is occurring can be identified by diagnostic means and


a determination made of the need to shut down and correct.

It is necessary to differentiate between a gradual reduction in


output and the gradual wearing of seal strips. This is particularly true
after returning to service from maintenance, and is a normal deteri-
oration caused by deposits on the blade elements. Step changes are
difficult to identify, as damage will often occur during transients and
the change may not be noted when the same conditions are reestab-
lished. Operating staff, particularly those with on-line monitoring
capabilities, best do this type of monitoring.

Increase in unit output at a constant valve position. Such a con-


dition normally occurs because of damage in the control stage noz-
zle block. This causes a change in wheel case and stage pressure
throughout the steam path. The nozzle plate will have sustained
damage and admits more steam because of an increased flow area.
If this increase is gradual it is due to some form of erosive damage
requiring refurbishment at the next available outage. However, if the
change is a step change, then the damage is sudden and should be
examined immediately.

Decrease in efficiency using enthalpy drop methods. If a utility


performing an enthalpy drop test (EDT) on a regular basis finds a sig-
nificant change in section state line efficiency, this normally indi-
cates some condition causing a disruption of flow and energy con-
version efficiency. Mechanical failure or vane deformation com-
monly causes the change of efficiency and should be considered an
indicator of existing problems.

Again, only step changes are significant. As EDTs may be run


only on an annual or six-month basis, a long time may pass before
such a condition is determined to be present by testing methods.
Station instrumentation used to monitor state line efficiency is usu-
ally not sensitive enough to note changes with sufficient reliability so
a unit could be opened on the basis of such results. Other forms of

10
Consideration of a Turbine Steam Path Maintenance Strategy

analysis would be required to confirm or deny the possibility of


damage.

Changes in steam extraction pressures. Blade system damage


will cause a redistribution of steam pressure throughout the steam
path. These changes may be minor, but if the unit is returned to sta-
ble and known operating conditions the changes may be determined
to exist.

The indicators discussed above do not represent an exhaustive


list, and in many units, operators and maintenance staff become suf-
ficiently familiar with individual units on their system to be aware of
unit peculiarities. Under these conditions, operating staff can often
act in an anticipatory manner and recognize characteristic changes
in the units. This ability represents a valuable skill to the owners, and
is only obtained from considerable experience with particular units.

System requirements
Forced outages mandate that owners arrange to meet system
requirements by obtaining power from alternative units. This alterna-
tive power can come from several sources. These include:

Line units on spinning reserve or stand-by. These units may


be at or close to the facility but the fact that they are spinning
reserve indicates their operation is less cost effective than the
units forced from service

Older, less-efficient units on the system. This may cause a


time delay but will allow system requirements to be met

Delaying any planned or scheduled outage on other units on


the system

Purchasing power from another utility connected to the sys-


tem. This can be expensive, but may be necessary in certain

11
Turbine Steam Path Maintenance and RepairVolume One

circumstances to meet requirements. Industrial installations


normally have ties to the local utilities, but this can represent
a significant economic penalty, particularly if boilers have to
be kept running to meet process steam demands

Adjustment of power factors and distribution voltage. There


is a limit as to how much can be achieved by this method.
On a large system this can make a difference but there is a
possibility of damage to certain electrical appliances, partic-
ularly from the lowering of voltage

In the event of a forced outage the owner must decide which


option offers the best alternative for meeting system demands. An
initial decision may be made in terms of returning service to cus-
tomers, but a long-term solution may require a change of supply to
meet the most cost-effective means of meeting demands over the
period of the forced outage.

OUTAGE SCHEDULING
A unit in service requires regular inspection or preventative
maintenance outages so it can be examined and corrective action
taken to allow it to remain in service in an acceptable condition.
There are two basic approaches to scheduling outages that provide
sufficient time to allow examination, the correction of known
defects, and the detection of any deterioration that will require atten-
tion at the next available outage:

Shut down the unit and expose sections and valves to make
an examination and determine required repairs. This is prob-
ably the most common approach, and is used by the major-
ity of large systems. This requires the entire unit to be made

12
Consideration of a Turbine Steam Path Maintenance Strategy

unavailable to the system for a period so all sections can be


worked on at the same time

Shut down sections selectively, opening one or possibly two


during a regular outage. The advantage here is that the
unavailable time is somewhat shorter and skills are con-
centrated on the opened sections. It also means that sections
prone to damage or significant deterioration can be exam-
ined more frequently

This dual approach has been found to be effective on systems


that have both summer and winter peak load demands combined
with relatively short periods in between, when the unit can be made
available for maintenance. It has also been used on large nuclear
units that must be shut down for six to seven weeks for refueling. At
that time one section of the unit is opened and made available for
replacements or refurbishment of damage.

INTERVAL BETWEEN
MAINTENANCE OUTAGES
Operators of steam turbine generator units must determine the
optimum, or acceptable interval between maintenance outages and
inspections. This interval between outages should be determined for
each unit within the system and must reflect not only operating and
repair costs, but also the consequences of not making such inspec-
tions and detecting faults in their early stages of development. One
manufacturer has determined that under certain conditions, and
with recognition of limitations, their unit rotors may operate for up
to 10 years between major re-inspections.

13
Turbine Steam Path Maintenance and RepairVolume One

Apart from the technical factors, the following issues should be


considered in establishing the period between inspection outages:

The fuel costs of the unit being removed from the system for
inspection and the unit(s) started to meet load requirements,
or the cost of the replacement power that will be purchased

The anticipated load factor of the unit during the outage peri-
od and its output rating

The differential heat rates of the unit being inspected and the
units being started to meet load requirements

The generic or historic forced outage rate of the unit

The anticipated maintenance period

Known labor costs and the anticipated cost escalation of


replacement parts

The anticipated improvement in unit heat rate that occurs as


a consequence of the outage

The reserve capacity on the system and the level of risk the
owner is prepared to accept in continuing to operate if the
outage is foregone

Problems discovered on similar units owned and operated by


the owning and other utilities

These factors influence the cost of an outage and the period for
which a unit should operate between major inspections.

The ideal situation for scheduling outages would be for a unit to


be open very infrequently; for no damage (or a minimum amount of
damage and deterioration) to be found at the outage, and then for the
unit to be successfully returned to service. In fact, units are opened
periodically on a life cycle basisusually between four to eight

14
Consideration of a Turbine Steam Path Maintenance Strategy

yearswith the possibility of this operating period being extended as


system reserve declines.

The actual opening time between major outages is dependent


upon various factors including:

the manufacturers recommendations. These are most often


based upon observations of a large population of similar
units and the experience established with these units after
different periods of operation

known problems within the unit. These can be related to


both supplier and user-induced situations. Design weakness-
es require opening the units more frequently until a perma-
nent solution to a particular problem is fully identified and
solved

the manner in which the unit has been operated since the last
outage. This is a factor over which the supplier has no con-
trol and, to a degree, the operator has only limited influence

Factors that influence the determination of when the unit should


be opened include:

the number and type of starts since the last outage (including
very hot, hot, warm, and cold starts)

any temperature transients experienced since the last outage

any excessive overspeed transients, and their duration since


the last outage

A new unit represents the latest technology available to the sys-


tem, and it is normally operated at maximum capacity. Unfortun-
ately, as the unit ages, it becomes less important to the system, and
component material properties will tend to degrade. The unit is then
normally subjected to more cyclic operation, which consumes com-
ponent life much faster.

15
Turbine Steam Path Maintenance and RepairVolume One

There is at present no absolute method of defining when a unit


should be removed from service for maintenance. On those occa-
sions when operators know it should be shut down, some operators
cannot because of system or process demands. Under these circum-
stances, the unit is operating at risk. If the operation cannot be avoid-
ed entirely, it should be minimized and the transients to which it is
exposed limited.

Other factors and considerations influencing the period of oper-


ation before a unit is removed from service for a maintenance out-
age are discussed later in this chapter.

Equivalent operating hours


A suitable method of assisting in opening decisions is to accu-
mulate the equivalent operating hours (EOH) for any service period
after the unit is returned to service from a major outage. In account-
ing for this period of operation, the EOH factors effect start ups and
can be made to include the effects of some transient operation. A for-
mulation developed by one major manufacturer provides the fol-
lowing equation that is applicable to their units, but with minor
modification can be made to suit any unit:
EOHs = (Nc x TC) + (NW x TW) + (NH x TH) + (NV x TV)

where:

EOHs = Equivalent operating hours due to starts


Nc = The number of cold starts
Nw = The number of warm starts
Nh = The number of hot starts
Nv = The number of very hot starts
Tc = The weighing factor for cold starts
Tw = The weighing factor for warm starts
Th = The weighing factor for hot starts
Tv = The weighing factor for very hot starts

16
Consideration of a Turbine Steam Path Maintenance Strategy

These time equivalent notations for starting the unit are added
to the normal operating hours (NOH), which are included irrespec-
tive of the load generated by the unit. When the total of normal oper-
ation plus start up reaches some predetermined value, the unit is due
for an inspection/maintenance outage. Therefore:
EOH = NOH + EOHs
These equations can be modified to include terms accounting for
both overspeed and temperature transients. A more difficult determi-
nation is the number of EOH that should form the basis of the times
at which the unit should be shut down for inspection. It is suggested
that an EOH of 30,000 to 35,000 be considered between major out-
ages. The initial and reheat temperature and the operators experi-
ences should modify this value with the unit.

THE INSPECTION/
MAINTENANCE OUTAGE
A principal purpose of the steam path maintenance inspection is
to detect potential problems at an early stage. If this is not done,
minor deterioration could progress to the extent that a forced outage
or excessive loss in unit efficiency could occur. During such an
inspection, parts can be examined visually for indications of failure,
wear, or distortion. Non-destructive tests can be applied to critical
components determining if their ability to continue performing satis-
factorily has deteriorated.

The satisfactory operation of a steam turbine within a utility is


influenced to a large extent by the manner and expertise with which
it is operated and maintained. For maintenance to be cost effective it
must be planned. When signs of distress, excessive wear, misalign-

17
Turbine Steam Path Maintenance and RepairVolume One

ment, or component deterioration are detected, the need for correc-


tive action must be considered. This corrective action should ensure
the situation does not deteriorate further, or to the extent the unit is
placed into a forced outage status, is severely load limited, or suffers
an unacceptably high degree of degradation in efficiency.

The cost of forced outages can be extremely high. In the case of


large utility units this often requires that older, less efficient units be
started to meet system demands or that units employing an alternate,
more expensive fuel be employed. The most severe situation for util-
ity systems is when reserve capacity does not exist and replacement
power cannot be purchased. At that time a blackout or brownout sit-
uation occurs.

In addition to routine maintenance, monitoring, and care activi-


ties, there are surveillance actions associated with a planned main-
tenance outage. In such instances, upper half covers are removed to
expose the steam path for examination. This opening of the unit
allows repairs to be made or replacement parts to be installed. At this
time, necessary skills and any special purpose tools that are required
can be made available.

Such a maintenance opening also allows replacement parts to be


ordered. This is done when an evaluation of any found operational
nonconformance is judged to place the unit at risk. An evaluation
of the nonconformance will normally indicate whether a unit can
safely be returned to service, or if some temporary remedial action
is required so the situation can be corrected before restarting.

18
Consideration of a Turbine Steam Path Maintenance Strategy

THE AVAILABLE
CORRECTIVE OPTIONS
The type of deterioration or damage to which a component or
area is subjected varies from unit to unit. It depends upon a variety
of causes that exist, or are present as a function of the components
design details and unit operating mode. These various causes can
normally account for the degree of concern the maintenance staff
expresses and they influence the different dispositions that will be
developed regarding the nonconformances.

In many instances of damage or failure there is rarely any best


or perfect solution. Instead, compromises will have to be made
based on the availability of replacement parts, overall costs, the
requirements for special tools, and the skills and time available to take
corrective action. The selected disposition will also be influenced by
the level of risk associated with each possible solution in any situa-
tion. The selected action should provide the best balance among cost,
risk, and how quickly the unit can be returned to service.

In order to rationalize the evaluation process, conditions and


maintenance decisions have to be defined. The most important follow:

A field found nonconformance


A nonconforming situation is considered to exist when the com-
ponents of the unit have changed or deteriorated to the extent the
design requirements are no longer present. A nonconformance in
any part of the steam path is considered to be present when there are
signs of structural failure, excessive wear, or any form of deteriora-
tion having the potential to adversely affect the performance of the
unit. When such a nonconforming condition is confirmed, it must be
viewed for both its short- and long-term effects (see chapter 11 for
manufacturing nonconformances).

19
Turbine Steam Path Maintenance and RepairVolume One

Fig. 1.8.1The Logic Review Process when a nonconforming condition is found


in the unit at maintenance inspection. The final decision of corrective action is
dependent upon many factors including the availability and delivery of replacement
parts. The maintenance engineer must evaluate the options and make a decision of
the best long term solution.

20
Consideration of a Turbine Steam Path Maintenance Strategy

When an inspection indicates a nonconforming condition, it


must be evaluated. The logic process of evaluation for performance
potential is shown in Figure 1.8.1. This figure outlines the options the
operator or maintenance engineer explores in deciding what correc-
tive action is most appropriate. An evaluation leads to one of four
alternate decisions. In certain circumstances, the decision is rela-
tively simple to make and in fact, is obvious. In others, options are
available, and a decision is made based on the probability of failure,
the possible cost of repair, and the ultimate consequences, including
the correction of consequential damage that results from not taking
corrective action.

The four basic decisions that can be reached are detailed below:

Scrap and replace. Such a decision is made when the situation has
deteriorated to the extent the component must be replaced, either
because it has failed, or returning it to service will jeopardize structur-
al integrity and safety of the unit. Often this is a self-evident decision
with little need for evaluation. At other times, this decision is reached
only after extensive review of options and possible consequences. In
this latter situation, it is judged that the risk associated with continued
use is too great, and the part must be scrapped and replaced, even if
this requires waiting for the delivery of replacement parts.

Under circumstances described as uncertain, a unit can be


operated at part load and/or with reduced steam conditions, if the
component suffering the nonconformance does not affect other parts
of the unit. Such a possibility must be evaluated for each noncon-
forming condition.

Repair. A repair corrects a nonconforming condition, but does


not re-establish original design characteristics within the element or
unit. It is often possible to make repairs to components, sufficient to
allow them to be returned to service. Depending upon the nature of
the nonconformance and of the repair, the affected component may
or may not ultimately require replacement.

21
Turbine Steam Path Maintenance and RepairVolume One

During the past decade there have been significant advances in


many repair and refurbishment techniques. This is particularly true in
cases involving welding, where new technology has made available
materials and techniques capable of extending the useful life of
many apparently failed or badly deteriorated components.

The technical requirements for performing such a repair are nor-


mally stringent. However, if they allow a unit to be returned to serv-
ice within a short period (rather than require an extended outage), or
until replacement parts become available, then the costs and mini-
mal change in risk levels associated with such repairs can often be
justified.

Again, the repair decision is normally made after a review of the


nonconformance, an evaluation of the possible repair procedures,
and the level of risk involved.

Rework (refurbishment). A refurbished component is considered


to be returned to its design condition (or better). The decision to
refurbish a nonconforming condition is, in some respects, similar to
that of repair but implies complete conformance with the original
design conditions.

In a number of situations, reworking involves reforming the exist-


ing material and may or may not require the addition of heat. In
many instances this is an easy decision to reach, particularly when
applied to stationary parts of the steam path. There can be an
extended time associated with major rework decisions, but many
techniques can now be undertaken without extending an outage.

Recent improvements in design and manufacturing technology


mean that for certain damage-condition refurbishments, components
that have the potential to perform at improved levels compare to the
original design. Also, there is a growing tendency among owners and
operators to repair or refurbish components once they have been
removed from the unit and replaced. The repaired/refurbished com-

22
Consideration of a Turbine Steam Path Maintenance Strategy

ponents can then be carried as inventory spares. If the component is


one that deteriorates during operation, the two elements can be
alternated at each outage. This is an attractive proposition, particu-
larly if there are a number of units in a plant or system utilizing these
same components.

Accept-as-is. This decision permits a component to be returned


to service with no effort made to correct the nonconformance. There
are three criteria involved in deciding upon this course of action:

There is little need to make any corrections. To do so will add


no more than marginal improvements to unit performance

The cost of replacing, repairing, or refurbishing cannot be


justified. This extends to the unit and the degree of deterio-
ration present as rework or repair could often increase risk to
performance level

There is insufficient time to take corrective action. Correction,


repair, or refurbishment would extend the outage period
beyond an acceptable time frame

This decision is often a judgment call based on the experience of


the engineer, and can only be made with awareness of any risks
involved. Such a decision should not be made as a desperation
measure. The risks, if any, should be fully evaluated and the options
considered, from an extended outage to operation and the probabil-
ity of failure.

An accept-as-is decision can often be made (again, being aware


of the risks) while replacement parts are obtained.

The decision or evaluation process can be a complex one.


Occasionally the solution is self-evident, i.e., when partial failure has
occurred or excessive damage exists. The most difficult decisions are
those related to suspected damage or deterioration, and those for
which it is difficult to determine the cause. In instances of uncertainty,

23
Turbine Steam Path Maintenance and RepairVolume One

mature judgment is required, together with knowledge of the operating


history of the unit. This knowledge can help in evaluating an existing
situation and can also provide more confidence in the selection of the
final disposition.

The availability of replacement parts, special skills, and tools will


often influence which decision is reached. Care must be exercised to
ensure that availability or non-availability of replacement parts does
not force the owner/operator into a decision that will ultimately be
more expensive and increase the overall risk to an unacceptable
level. Often alternate solutions are available that may degrade a
units rating or impose other restrictions, but are ultimately more
acceptable short-term, while the owner/operator arranges for a more
palatable long-term solution.

A common, logical, and recommended approach to mainte-


nance decisions and repair dispositions is to:

consider the alternate actions in terms of the design require-


ments of the affected components

evaluate each possible solution in terms of the departures


from the design-expressed requirements they represent

Many repair or accept-as-is dispositions will have only a minor


effect on unit performance and can be readily accepted. Other
repairs can be accepted if the unit can be operated in a compro-
mised condition, such as reduced steam conditions, or limited load
swings.

24
Consideration of a Turbine Steam Path Maintenance Strategy

DISTINCTION BETWEEN
CAUSES AND MECHANISMS
OF FAILURE
When a unit is openedeither in a planned outage or a forced
outage situationthere is the possibility that a failed, damaged, or
deteriorated condition will be found. It is necessary then for the
owner and the maintenance and operating staff to analyze the situa-
tion and select a corrective action plan.

A first step in discovering the existence of such a situation is to


identify the cause. This is of particular importance when a new or
relatively new unit is examined, as this condition can be sympto-
matic of a design fault that must be corrected. A second step is to
identify the mechanism of failure.

It is important to recognize at the beginning of such an investi-


gation, that the cause and mechanism causing the deterioration
are not the same things. The failure mechanism is the form of mate-
rial deterioration that has occurred and ultimately consumes the life
of the material or component, often resulting in material rupture or
severe deformation. The cause is that feature of the operation or
structure of the unit that has introduced the environment within
which the mechanism can initiate and develop.

The causes of failure


When a mechanical component fails, there is obviously a cause.
Failures do not occur if they are not initiated by some condition
existing within the unit. The possible causes are shown in Figure
1.9.1 and examined here.

Poor or inadequate design. Some possible causes of design-ini-


tiated failures are shown in Figure 1.9.2.

25
Turbine Steam Path Maintenance and RepairVolume One

Mechanical Failure

Mechanical Failure

Failure
Mechanism
Failure
Cause Mechanism
Cause

Design
Design Manufacture
Manufacture Installation
Installation Operation
Operation Maintenance
Maintenance

Possible Failure
Possible Mechanisms:
Failure Mechanisms:
HighHigh
Cycle Fatigue,
Cycle Fatigue,LowLowCycle
CycleFatigue, MechanicalParticle
Fatigue, Mechanical Particle Impact,
Impact,
Water
Water Impact
Impact Erosion,
Erosion, SolidParticle
Solid Particle Erosion,
Erosion, Water
WaterWashing
Washing Erosion,
Erosion,
Creep, Fretting or Stress Corrosion Effects.
Creep, Fretting or Stress Corrosion Effects.

Figure 1.9.1
Fig. 1.9.1The major causes of failure in the steam path, together
The major causes of failure in the steam path, together with the
with the principle failureprinciple
mechanisms.
failure mechanisms.

MECHANICAL FAILURE

CAUSE

DESIGN Manufacture. Installation. Operation. Maintenance.

Research and Calculation Specification Manufacturing


Development System Preparation Support

* Correct data * Incorrect data used * Incorrect transfer of * Poor review of


not established. to establish design. data to drawings and nonconformances.
* Misinterpret * Misinterpret specifications * Failure to inspect
Results. calculated values. *Poor choice of correctly.
*Poor or Inadequate * Poor use of computer tolerances. * Poor inspection
Equipment programs. * Unclear Definition of techniques and
* Poor selection of requirements. standards.
available component
options.
* Inadequate use of
existing research

Fig. 1.9.2The most likely influences Figure


from design
1.9.2 error which could contribute to
component failure.
The most likely influences from design error which could contribute to component failure.

26
Consideration of a Turbine Steam Path Maintenance Strategy

The design process produces a manufacturing specification that


identifies dimensional requirements, materials, and process require-
ments sufficient to make the unit suitable for its intended applica-
tion. To achieve this specification the design process employs the
results of research, development, and experimentation, all of which
are initiated by the need to identify and solve potential problems that
could arise with the performance of the unit.

In all large design organizations, qualified procedures exist for


both the thermodynamic and mechanical design procedures. These
are well understood by the industry, and applied by the individual
engineers undertaking the design tasks. There are situations, howev-
er, in which the extrapolation of existing designs, errors in calcula-
tions, or the incorrect specification of materials will cause deteriora-
tion in a unit sooner than normally expected.

Errors caused by the design process can correctly be attributed to


faulty design specification or technology application and implemen-
tation. With the use of computerized design techniques, many of
these errors, and the opportunity for such errors to occur, have been
lowered.

A basic premise of the design process is to define a component


that will operate without failure for a minimum of 200,000 hours
(about 30 years of normal operation). This philosophy is applied to
major components. It is recognized that other components are con-
sidered to be consumable and will be replaced on a regular basis
to maintain both efficiency and availability at acceptable levels. For
those components designed to achieve 200,000 hours of operating
life, it is assumed their units are operated as intended and that oper-
ating parameters, as defined by the designer, are maintained. In
defining these operating limitations, some or all of the following
restrictions may be addressed:

Initial steam pressure rate of variation and total hours of over-


pressure operation

27
Turbine Steam Path Maintenance and RepairVolume One

Initial and reheat steam temperature and hours of over-tem-


perature operation

Initial and reheat steam temperature; rate of variation

Condenser backpressure

Unit speed and speed swing limitations

Overspeed testing recommendations

The removal of feed water heaters

The rates at which the unit should be started, loaded, and


shut down

Normally these limitations are addressed in terms of the number


of hours of operation under off design conditionsthe number,
frequency, and severity with which the unit could experience various
transient conditions (recognizing many of these conditions are diffi-
cult or impossible to control). These limitations vary from manufac-
turer to manufacturer and reflect to some degree each manufactur-
ers operating experience under such conditions. They may also
reflect the basic design philosophy and, to some extent, the level of
conservativeness in both the design and operating philosophy.

The design engineering function is responsible for unit and com-


ponent quality, as well as the definition of the tolerances to which
the unit and its components will be manufactured and tested to
achieve the level of required performance. Therefore, when noncon-
forming conditions result within a manufacturing department, the
design engineer must evaluate them, rule on their acceptability, and
decree the corrective action required.

In the event the design engineer accepts a situation or condition


not within his defined tolerances, and the unit is forced from service
or a planned outage has to be extended to correct such a situation,

28
Consideration of a Turbine Steam Path Maintenance Strategy

this is considered a design initiated failure. The root cause is poor


design.

Poor quality manufacture. It is the responsibility of the manu-


facturing department to convert an engineering specification to a fin-
ished product. The manufacturing department has no responsibility
for establishing the quality level of the individual componentsonly
to ensure that quality is met (as defined by the design department).
Quality is defined in a series of documents from drawings to
process and material specifications. Normally each of these docu-
ments are supplied with production and application tolerances
clearly defined, either as limitations placed on the individual docu-
ments or as standards established and published within the manu-
facturing department.

It is obvious, therefore, that there exists considerable opportuni-


ty for manufacturing errors to occur. To help prevent or minimize
them, the engineering department employs quality control and qual-
ity assurance methodology. (The industry is fortunate that there are
not a greater number of faults and failures occurring due to inade-
quate manufacture.) These quality programs are a management-
imposed discipline. However, manufacture is still a human opera-
tion and there are still a number of hand operations and judgmental
situations associated with manufacture. Faults can and do occur and
are incorporated into a unit.

The major areas of manufacturing activity where faults can occur


are shown in Figure 1.9.3.

Component assembly is an integral part of the manufacturing


process. It carries with it the possibility of residual stresses which,
compounded by operating stresses, can cause premature failure of
components. Unfortunately, such failures are difficult to trace and
cannot be anticipated, since the deformed or over-stressed element, or
the lack of inspection, is often hidden by the assembled components.

29
Turbine Steam Path Maintenance and RepairVolume One

In turbine manufacture, many operations are termed special by


quality definition. The term indicates that at the completion of the
process, there is little or no opportunity to check conformance with-
out destructive testing or significant degradation of the final product.
Such processes are often associated with heat fusion or assembly. In
these instances, quality assurance involves calibrating the process to
ensure conformance with the process. This is done by the detailed
examination of samples to achieve the correct product, including
destructive testing and metallurgical examination. However, during
the application of such a process it is possible for calibration to slip,
and so these processes often represent manufacturing areas where
nonconforming products can be incorporated into the unit if careful
control is not exercised.

MECHANICAL FAILURE

CAUSE

Design. MANUFACTURE. Installation. Operation. Maintenance.

Metal Process Metal Component


Production. Control Forming Assembly.
Testing

* Mechanical * Heating. * Cutting. * Incomplete Testing


* Misalignment
Properties. * Fusion. * Casting. * Inadequate
* Interference.
* Chemical * Heat * Forging. Procedures.
* Pre-stress.
Properties. Treatment. * Inadequate
* Overstress.
* Heat Instrumentation.
* Incorrect fit.
Treatment. * Instrumentation
* Microstructure. not calibrated.

Fig. 1.9.3The most likely influences from manufacturing


Figure 1.9.3 error which could contribute
toThe
component
most likelyfailure.
influences from manufacturing error which could contribute to component failure.

30
Consideration of a Turbine Steam Path Maintenance Strategy

Poor quality installation. Assembly and installation on-site is a


complex operation that includes the installation of individual com-
ponents and sub-assemblies. This should also be considered to
include commissioning, which is normally an integral part of the tur-
bine supply process. If these activities are not performed in a correct,
controlled manner they will often result in poor operating character-
istics and the imposition of undue stresses on the component parts of
the unit. These additional strains can include residual stresses leading
to premature failure within the components of the unit if not detect-
ed and corrected. Some possible causes of installation- and total
assembly requirement-initiated failures are shown in Figure 1.9.4.

Site assembly requires that multiple section units be aligned in


vertical and horizontal locations and all clearances are at design
specified values to achieve the correct relationship between the

MECHANICAL FAILURE

CAUSE

Design. Manufacture. INSTALLATION. Operation. Maintenance.

Shop Shop Field Start up


Erection Testing Erection

* Incorrect assembly * The use of incorrect, * Failure to achieve * The use of incorrect
and sub-assemblies. or inadequate procedures. design specified procedures.
* Incorrect fits in the * Inadequate test alignment * Misapplication of
assembled procedures. * Failure to achieve procedures.
components. * The use of poor testing clearances at * Inadequate testing and
* Incorrect component techniques. critical positions. setting of trips and
alignment. * Failure to correct found control mechanisms.
discrepancies.

Fig. 1.9.4The most likely influences Figure


from poor 1.9.4
installation which could contribute to
The most likely
component influences from poor installation which could contribute to component failure.
failure.

31
Turbine Steam Path Maintenance and RepairVolume One

rotating and stationary components during operation. Failure to


achieve the correct alignment will result in higher than expected lev-
els of vibration and possibly the imposition of alternating stresses on
the normal operating levels predicted by design. Poor control of
alignment can also result in operating rubs that have the potential
to degrade efficiency and reliability.

Control gear and instrumentation must also be installed and cal-


ibrated to ensure required levels of control and protection.

Poor or incorrect operating practices. A common cause of com-


ponent deterioration or failure is the application of incorrect operat-
ing procedures, or violation of the operating parameters. This is par-
ticularly relevant in temperature variations. In many instances the unit
operator may have little control over such variations and may have to
respond to the requirements of the system controller so customer
demands can be met. However, such conditions required by the sys-
tem controller must still be considered faulty operating procedures.

It is normal for the design engineer to place certain limitations or


restrictions on the manner in which the unit is operated. These limi-
tations are not arbitrary, and are intended to minimize stresses in all
forms to which the unit components are subjected. Operating out-
side these limitations will consume the life of the elements and can
lower the predicted life. The rate at which the unit or component life
is consumed also has an effect on the unit reliability.

Figure 1.9.5 shows some of the operating actions that can result
in component failure, life consumption, or deterioration. It is clear
the operation of other system components can also influence the
steam turbine. The operator, therefore, has a responsibility to ensure,
as far as possible, other pieces of equipment do not impact adverse-
ly on the performance of the steam turbine.

Poor maintenance procedures. Steam turbine maintenance falls


into two categories:

32
Consideration of a Turbine Steam Path Maintenance Strategy

MECHANICAL FAILURE

CAUSE

Design. Manufacture. Installation. OPERATION. Maintenance.

Operator Failure of
Thermal Excessive
Error Other System
Transients Testing
Equipment

* Poor turbine- * Incorrect start-up and * Debris carried into * Excessive valve
Boiler coordination. shut down methods. steam path. testing.
* Excessive thermal * Excessive start-up * Poor steam condition * Excessive overspeed
change rates. rates. control. testing.
* Excessive loading * Poor synchronizing * Excessive steam passed
rates. procedures. through turbine.
* Lack of system control.

Fig. 1.9.5The most likely influences from poor


Figure operating practice which could con-
1.9.5
The most likely
tribute to component influences from poor operating practice which could contribute
failure.
to component failure.

Routine maintenance activities, i.e., valve testing, oil level


maintenance, greasing slides and supports, etc.

Maintenance performed to upgrade or maintain an accept-


able condition when a unit is opened up

In terms of unit performance, this second aspect of maintenance


is related to component evaluation and the correction of noncon-
forming conditions. This tends to be more critical in terms of discov-
ered existing and potential failure, but the routine maintenance, if
neglected, can also have serious and long consequences for the unit.

Figure 1.9.6 lists some maintenance activities that can influence


the performance of the steam turbine.

33
Turbine Steam Path Maintenance and RepairVolume One

MECHANICAL FAILURE

CAUSE

Design. Manufacture. Installation. Operation. MAINTENANCE.

Normal Outage Repair and


Maintenance Determined Refurbishment
Activities Maintenance. Activities

* Failure to undertake * Failure to correct a * Use of unqualified skills


Recommended found nonconformance. to perform repairs.
Maintenance. * Correction by inadequate * Use of incorrect tools to
* Incorrect Procedure. perform repairs.
Maintenance * Correction by inadequate * Misapplication of
Procedures. methods. procedures.
* Use of incorrect
materials and
approved substances.

Fig. 1.9.6The most likely influences Figure from poor


1.9.6maintenance procedures which could
contributeThe
tomost likely influences
component from poor maintenance procedures which could contribute
failure.
to component failure.

The examination of a failure


Considering the causes listed above, when a failure or deteriorat-
ing condition is found to exist, it is important to identify the actual
cause and, to the greatest extent possible, establish if it was the result
of design, manufacture, installation, operation, or maintenance. This
can become a difficult and often impossible task. There are no rules
or guidelines that can be applied. It often takes considerable investi-
gation to identify both the cause and the initiating condition.

Often, if the initiating condition can be identified (i.e., inferior or


inadequate material, poor process control, overheating of the com-
ponents, etc.), it becomes considerably easier to establish which of
the five causes discussed above is the initiating condition.

34
Consideration of a Turbine Steam Path Maintenance Strategy

The cause of a nonconforming situation within a uniteven


when the mechanism is well defined and agreed uponis often a
contentious matter among the parties involved in correcting it.

In fact, in some instances, it becomes impossible to isolate a sin-


gle cause beyond a reasonable doubt. It should be remembered that
the primary reason for identifying the cause is not to apportion
blame (except in the case of new units or components where war-
ranty considerations exist), but to allow the manufacturer and plant
staff to initiate corrective actions that will prevent or minimize the
possibility of its reoccurrence.

Figure 1.9.7 is a logic diagram for the investigation and identifi-


cation of causes and mechanisms for a damaged or deteriorated
condition.

This investigative process is shown to comprise two phases. The


first might be termed a field phasewhile all material and infor-
mation is at the site, work is undertaken to allow the condition to be
corrected and the unit returned to service. There are many occasions
when this will suffice and no further investigation is required.

However, if this field phase does not identify the cause and/or
mechanism, then the second is initiatednormally after the unit has
been returned to service. There are instances when a condition can-
not be corrected and restorative action taken until this laboratory
phase is completepossibly after some redesign and/or corrective
manufacture has been undertaken. This need is determined by the
extent and severity of the failure.

Initiating and driving mechanisms


When attempting to analyze a failure, it is necessary to identify
the cause as far as possible. It is also necessary to identify the mech-
anisms involved in the failure. This determination can be difficult.
For a well-designed and well-operated unit, the stress levels and

35
Turbine Steam Path Maintenance and RepairVolume One

FAILURE OR SIGNIFICANT
INITIAL PHASE
DETERIORATION DISCOVERED
(Field Phase).

(a)

Determine Location Record Condition. Review Operating


and Extent of damage. Procedures.
* Photographs. * Starting and Loading
Is this an isolated condition, or is the condition Macro and general Procedures.
occurring in more than one location within the arrangement and location. * Thermal Transients.
stage and are more than one location or stages * Measurements * Vibration Characteristics.
involved? * Sketches. * Maintenance Practice.

Sample Collection.

Examine Initiation
point.
Debris or any detached Surface deposits
pieces
available

Evidence of Damage
and Surface Hypothesize on Initiating Hypothesize on
Corrosion. Mechanism Driving Mechanism
Erosion.
Fretting.
Water Damage.
There may be sufficient evidence available from a field
Impact Damage.
examination that further detailed analysis is not required,
High Cycle Fatigue.
and recommendations for the unit can be made.
Other Mechanisms.

EXAMINATION OR
EVALUATION If Field Observations
(Laboratory Phase) cannot establish causes

(b)
Collect:
Failure Samples Preliminary
Photographs. Visual Analysis of
Surface Samples. deposits.
Operating Records.

Finite Element Microscopic Examination


Stress and Frequency of failure surface.
Analysis

Evaluate results and


Frequency Testing
Establish Cause of
and Analysis.
failure.

Recommend:
New Materials.
New mounting Procedure.
New Operating Procedures.
Change to Maintenance Procedures.
Change of water chemistry.

Fig. 1.9.7Two phases of the damage Figure 1.9.7 process for identifying and
evaluation
Two phases of the damage evaluation process for identifying and quantifying a field found failure.
quantifying a field found failure.

36
Consideration of a Turbine Steam Path Maintenance Strategy

operating procedures should guard the various components against


premature failure. Therefore, when a failure or deteriorated condi-
tion is found, it is important to identify the mechanisms producing
this condition.

In making this determination it is normally necessary to identify


two separate mechanisms. The first is that which initiated the condi-
tionwhich can be any of a number, as discussed below. The sec-
ond is the mechanism that is driving the condition, often to failure.
In the majority of situations, this mechanism is high-cycle fatigue,
ultimately producing material rupture at stress levels well below
those able to produce failure if an initiating condition or fault had
not been induced by some other stress-concentrating condition (i.e.,
the initiating mechanism).

The mechanisms of damage and deterioration


The mechanisms of failure are the normal deteriorating process-
es such as fatigue and corrosion, which can occur in the majority of
mechanical components if the conditions are right. Many of these
mechanisms are relatively easy to recognize. They are known to
occur as a result of the environment in which components of the unit
operate and the loads to which they are subjected. In terms of the
steam turbine the major mechanisms are:

High-cycle fatigue (HCF). Possibly the most damaging effects


present in the turbine steam path are those associated with HCF.
HCF is a mechanism that is a primary or initiating cause of failure in
many components, both rotating and stationary. It is also the major
damaging mechanism that will drive damage initiated by some other
mechanism, forcing the situation to propagate to destructive levels.

Cyclic effects (or impulses) are generated in the flowing steam due
to a number of effects (discussed in chapter 5). The frequency of the
impulse is dependent upon the nature and source of the excitation,

37
Turbine Steam Path Maintenance and RepairVolume One

and the magnitude is dependent upon the flow characteristics at that


location in the flow path. (Table 5.10.1 of chapter 5 lists various excit-
ing sources within the steam path.) It is not necessary for the impulse
magnitude to be high for failure to occur. Also, if the frequency of the
impulse is near one of the natural frequencies of the damaged com-
ponent, it requires only a relatively low magnitude impulse to cause
failure.

A typical surface showing the effects of HCF is shown in Figure


1.9.8. The beach marks characteristic of HCF can be clearly seen.
In fact, the true HCF surface shows a fine, uninterrupted crystalline
structure, and the final rupture is in the form of a tensile or torsional
overload failure, depending upon the form and type of stress causing
final separation. The beach marks are normally present and represent
transient conditions causing a fracture arrest or change in the crack
propagation rate. This can be a massive progression of the crack at
the tip of its current position.

TAPPS

Fig. 1.9.8The striations or beach marks on the vane of an L-1 blade, and the mating tip
section recovered from the internals of the turbine.

38
Consideration of a Turbine Steam Path Maintenance Strategy

Section Harmonic Typical Sources

High Pressure High per rev Nozzle tolerance limits


Section (40x) Upstream wake degenera-
tion Structural turbulence

Intermediate Nozzle passing Nozzle wakes


Pressure frequency (NPF)
Section 2 x NPF Diaphragm harmonics
3 x NPF Diaphragm harmonics

Low One per rev Relative displacement


Pressure nozzles to blades
Section 2 per rev Diaphragm joints
Multiple/rev Structural supports
in flow path
Medium/rev Diaphragm harmonics
Aeroelastic disturbances
High/rev Nozzle turbulence
harmonics
Upstream wake
degeneration
Structural turbulence

Table 1.9.1Steam Path Exciting Forces

The true HCF surface has no indication of mechanical deforma-


tion, and if a surface were to fail entirely by HCF, the mating surfaces
would fit together perfectly.

Low-cycle fatigue (LCF). Low-cycle fatigue is distinct from the


HCF phenomena in several ways. These are best summarized as:

the number of cycles required to cause failure are signifi-


cantly less than for HCFthe time to failure being a function
of the magnitude of the stress levels

the number of stress cycles to cause failure are significantly


less than for HCF. An arbitrary demarcation is often taken to
be 100,000 cycles. However, this is simply a convenient

39
Turbine Steam Path Maintenance and RepairVolume One

means of definition. Surface appearance and other charac-


teristics provide a better delineation between the two

the individual stress levels developed at each cycle are nor-


mally considerably higher with the low-cycle effects

there is mechanical deformation at the failure surface, and


failure marks show individual evidence of rupture, i.e., the
stresses are large enough to cause plastic deformation at the
leading edge of the crack

The most common forms of LCF failure are those attributed to


thermal cycling, when large temperature changes occur causing
localized thermal expansion or contraction of the material surface.
These thermal movements introduce large internal constraining
forces within the material introducing stresses of a magnitude suffi-
cient to initiate and drive to rupture.

Figure 1.9.9 shows a normal LCF crack that originated at a sharp


corner within a high temperature casing. Such cracks can be
repaired normally by welding or by stitching.

Fig. 1.9.9A low cycle fatigue crack in a high pressure casing. This crack originating
at a point of high stress concentration adjacent to the nozzle plate.

40
Consideration of a Turbine Steam Path Maintenance Strategy

Creep deformation. Steels that are subjected to elevated temper-


atures for extensive periods exhibit changes in their microstructure.
This has a marked effect on the material creep characteristics. Under
conditions of high tensile stress, failures are normally trans-granular.
However, at elevated temperatures and with the application of ten-
sile stress, it is common for inter-granular failure to occur due to the
effects of creep. Creep is a granular slip mechanism within the
material that occurs at dislocations of the grain boundaries. This slip
causes a plastic distortion of the components that is not removed
when the stress is removed from the component.

On high temperature stages there are instances in which creep


deformation causes plastic deformation of the steam path compo-
nents. This is not a common type of mechanism, and occurs in only
the highest temperature elements. It is often difficult to detect with-
out the aid of detailed and repeated measurements. However, in
those instances where this phenomenon is present, it contributes to
the general degradation of a stage or component.

It is uncommon for elements to operate for periods of time after


they go into the tertiary region of creepand failas the damage is
normally obvious before this level of damage is reached. Figure
1.9.10 shows the notch block in a high-temperature stage in which
the centrifugal load of the blade is causing creep deformation of the
pin and wheel material. Such damage presents an increase of the
gap between the closing block and wheel. This type of gap is com-
mon but does not necessarily indicate that failure will occur. The
width of the gap should be monitored until a maximum stipulated
gap has been formed.

Solid-particle erosion (SPE). SPE is a damaging mechanism


found in many front-end stages of high-pressure and reheat sections.
SPE removes material from surfaces through a combination of goug-
ing and the impact of solid particles of hard scale exfoliated from
boiler tubes and carried into the steam path.

41
Turbine Steam Path Maintenance and RepairVolume One

Fig. 1.9.10The notch block from a high temperature stage


which has crept out radially under the influence of cen-
trifugal loading.

As scale enters the steam path, it impacts on and removes mate-


rial from the stationary and rotating blade elements, the sidewalls,
and the coverbands. It is also capable of removing coverband tenon
material, thereby weakening the attachment of the coverband to the
blades.

On stationary blade rows, the major effect is to remove material


from the vane discharge edge, causing thinning and an increase in

42
Consideration of a Turbine Steam Path Maintenance Strategy

Fig. 1.9.11Solid particle erosion showing material loss from a fixed blade discharge
edge. In this stage the material loss is from previously weld repaired elements, and the
loss has been severe. This material loss modifies the total discharge area from the stage.

the stage discharge area and angle (Fig. 1.9.11). The sidewalls can
also suffer material removal, normally at the outer sidewall. The inlet
to the stationary blades can also be affected by the gouging action of
the particles (Fig. 1.9.12).

On the rotating blades, material is removed from the inlet edge,


modifying the vane section and inlet angle (Fig. 1.9.13) and can
remove material from the discharge edges (Fig. 1.9.14). From a struc-
tural consideration, a more insidious material loss comes from the
pressure or concave face of the profile (Fig. 1.9.15). Such damage
should be looked for when damage of the form shown in Figure
1.9.14 is evident. This is because many stages subject to this erosive
phenomenon have an integral coverband. This material loss repre-
sents an undercutting of the integral coverband that will weaken the
outer section of the blade and modify its vibratory characteristics.
For stages without integral coverbands, there can be an undercutting
of the tenons.

43
Turbine Steam Path Maintenance and RepairVolume One

Fig. 1.9.12Material removed from the inlet nose of a


stationary blade row by solid particle erosion.

Fig. 1.9.13Material removed from the inlet edge of a rotating blade row by solid
particle erosion.

44
Consideration of a Turbine Steam Path Maintenance Strategy

Fig. 1.9.14Material loss from the discharge edge


of a rotating blade row by solid particle erosion.

Corrosive effects. The ingress of corrosive ions into the steam


path is always potentially possible. These ions have the potential to
cause dramatic and expensive outages within the unit, and many
stages of the turbine can be affected by this condition. However, the
most seriously affected stages are usually those in which moisture
forms after expansion of the working fluid into the moisture region.
As this occurs, many of the corrosive ions precipitated from the
steam are deposited on the steam path surfaces and the deposited
moisture carries them into hideouts where they concentrate. At
these locations they become aggressive when the correct environ-
mental conditions occur.

45
Turbine Steam Path Maintenance and RepairVolume One

Fig. 1.9.15The effect of caving on the pressure


surface of a rotating blade.

There are several corrosive conditions having the potential to


cause damage or deterioration:

Stress corrosion cracking (SCC)For SCC to occur, three


conditions must be present in the elements: the presence of
aggressive ions, a tensile stress in the component, and the
material must be susceptible. The rate at which such cracks
initiate and propagate depends upon the environmental tem-
perature present to support the chemical reaction. In many
portions of the steam path, residual tensile stress existsboth
by design and also accidentallyas a consequence of man-

46
Consideration of a Turbine Steam Path Maintenance Strategy

ufacture. At these locations SCC can readily occur. The rate


at which corrosive cracks are generated and then propagate
depends upon the concentration of corrosive products and
the magnitude of the stress and local temperature

Corrosive pittingFigure 1.9.16 shows a rotor portion where


High-Lo seal castellation occurred. An environment of
geothermal steam with a high corrosive content has corrod-
ed these

Corrosion fatigueMany components are subject to alter-


nating stress of a magnitude that ultimately, fatigue con-
tributes to its deterioration. If a corrosive element further
contaminates the component, then the components ability
to resist failure is further impacted. Under these circum-
stances, failure will occur at either lower stress levels or in a
shorter period of time at the same stress. Figure 1.9.17 shows
a crack emanating from a tie-wire hole, where the initiating
mechanism of failure was found to be corrosion fatigue

Fig. 1.9.16Corrosive pitting on a rotor surface.

47
Turbine Steam Path Maintenance and RepairVolume One

Fig. 1.9.17A crack initiating at the tie wire hole.

This latter form of failure is distinct from SCC in that the applied
stress is cyclic and the failure is a fatigue-type mechanism. Such fail-
ures tend to initiate at those points where corrosive products can
concentrate and where a level of stress concentration causes high
local tensile effects. Such cracks can be either trans-granular or
inter-granular.

Moisture effects. Moisture forms in the working fluid due to con-


densation from the expanding steam (see chapter 3). Water droplets
that form are transported through the steam path by the parent
steam. Some of these droplets are ultimately deposited on the com-
ponent surfaces where they have the ability to flow across them and

48
Consideration of a Turbine Steam Path Maintenance Strategy

accumulate into larger droplets. These larger droplets can re-enter


the main steam flow and ultimately impact with the metal surfaces
of the steam path components, causing erosion. However, more than
one form of water damage can occur:

Moisture-impact erosionOnce water droplets have formed,


accumulate, and re-enter the steam path, they have the
potential to erode the blade material if they impact at a value
above a threshold velocity. This form of damage is common
on the majority of large last-stage rotating blades (Fig. 1.9.18)

Trailing edge erosionThis damage is caused on the dis-


charge edge of rotating blades by water introduced into the
steam path to cool the exhaust (Fig. 1.9.19). It occurs in high
stress areas and is capable of causing ruptures that initiate at
one of the erosion/cavitation grooves at the edge

Washing erosionWhen water flows at high velocities


across metallic surfaces there is always the possibility of sur-
face material loss. As this occurs, material grains break away
and cause a roughening of the surface. Figure 1.9.20 shows
the outer ring of a cast-iron diaphragm where material has
been removed by washing, causing grooves to be produced
emanating from the suction surface of the vanes

Wire drawing erosionWhen water exists at high-pressure


differentials across a joint, there is always the possibility this
moisture will be forced across the joint, and will remove sur-
face material. This loss is also sometimes called worming
(Fig. 1.9.21)

Water ingestionIf water is returned to the turbine steam


path as a slug this can cause various forms of damage from
blade rupture to rotor bending

49
Turbine Steam Path Maintenance and RepairVolume One

Fig. 1.9.18Moisture impact erosion.

Transient operating conditions. Severe temperature ramps are a


common damage phenomenon associated with transient operation.
Other considerations include overspeed transient, which occur on a
sudden load rejection or when testing the valve mechanisms. Each
of these mechanisms or situations has the potential to cause a dete-
rioration in the operating capability of the unit.

Fretting corrosion. When there is relative motion between sur-


faces that are in nominally tight contact, heat generation can result
from this movement. The movement may be of the sliding or

50
Consideration of a Turbine Steam Path Maintenance Strategy

impact type. In either case there is a generation of excessive heat


that removes surface material by causing localized oxidation. The
material removed has the characteristic red rust appearance.

Fig. 1.9.19Trailing edge erosion occurring on the last stage blades.

Fig. 1.9.20Washing erosion on the outer ring of a cast iron diaphragm.

51
Turbine Steam Path Maintenance and RepairVolume One

Fig. 1.9.21Worming erosion on the half joint of a diaphragm.

COMPONENT SUSCEPTIBILITY
FOR DETERIORATION
Should either the critical fixed or rotating components within the
steam turbine sustain damage, it could be sufficient to force the unit
from service. Such damage can be caused by various mechanisms
associated with operating phenomena resulting from the characteris-
tics of the working fluid, its pressure, temperature, moisture content,
and possible rate of change during operation (these were discussed

52
Consideration of a Turbine Steam Path Maintenance Strategy

in the previous section). Deterioration can also be related to the


stress levels within the components, or the result of matter carried
into the unit from the boiler or other parts of the system. Problems of
deterioration can also be the consequence of defects in the manu-
facture of the component.

In terms of the level of damage or deterioration with which the


maintenance staff should be concernedand possibly be monitor-
ingthe following categories of deterioration potential or suscepti-
bility can be ascribed to the various locations and components
within the steam path. While not exhaustive, they will provide some
general indication.

High susceptibility
Typically, these components include the last (L-0) stage rotating
blades and many L-1 or L-2 stages. These longer blades are tuned,
and the last stage blades are subject to variable pressure ratios and
damage as a consequence of their interaction with the moisture pres-
ent in the steam. This category would also include the high temper-
ature stages of the high pressure and reheat sections that operate on
steam returned from the boiler. In these stages, exfoliated scale may
be carried in with the steam, causing stationary and rotating blade
vane material loss. If the first stage of the high-pressure section is also
subject to the effects of partial arc admission, that can also cause
high impulse loads on the stationary and rotating blades, possibly
leading to high-cycle fatigue failures.

Wheels shrunk onto a central spindle can have a high potential


for damage due to corrosive action. Thisin terms of serious dam-
agecan be avoided to some extent by the stringent control of water
quality and careful examination at outages.

53
Turbine Steam Path Maintenance and RepairVolume One

Intermediate susceptibility
Typically, these components include those subject to corrosive
attack, such as the last minus one (L-1) stage in the low-pressure sec-
tion. Depending upon the pressure distribution, it could also include
the last minus two (L-2) stage. This group also includes stages with a
drilled hole in the vane for the admission of tie wires. The second
and third stages after admission or return of hot steam can also be
included because of degraded material mechanical properties at
high temperatures. Because pressure loads on the various stage com-
ponents can be high, creep deformation is a consideration. The pos-
sibility still exists with these front-end stages that hard scale from the
boiler will cause an erosive material loss.

The diaphragms of the high temperature stages can be subject to


creep, and rubbing should be considered as a potential. Nozzle
boxes in the high pressure section can be a cause for concern due to
solid-particle erosion and possibly other damage resulting from tem-
perature transients and debris carryover from the boiler, particularly
after boiler repairs.

Low susceptibility
This category includes all other components and stages in the
unit. The blade rows operate at substantially constant pressure ratios.
Operating temperatures are normally below the threshold tempera-
ture where material properties are reduced significantly, so compo-
nents are not normally subject to creep deformation.

The rotor wheels, in many stagesparticularly if there are pres-


sure balance holeshave a potential to cause problems, and fillet
radii can eventually be the initiation point of low-cycle fatigue
cracks. In wet stages, these wheels can be subject to moisture-impact
erosion under certain operating conditions, although the threshold
velocity at which such erosion normally occurs has not been
achieved.

54
Consideration of a Turbine Steam Path Maintenance Strategy

Categorization
In categorizing components, it must be remembered that many
units appear to have characteristics which make certain stages and
components within those stages more susceptible to damage and
deterioration. Therefore, categorization as given above should be
considered as only a guide, and should not be interpreted to mean
the low susceptibility elements will not suffer damage until after
the high and intermediate susceptibility elements have been
affected. Most owners will be aware of rogue stages and compo-
nents that appear to fail with monotonous regularity in an otherwise
acceptable unit.

While some elements may have a high susceptibility to failure or


deterioration, it is possible that in the event of failure, these can be
removed and the unit will continue to operate. Conversely, it is pos-
sible that low susceptibility elements must be replaced or refur-
bished when they do deteriorate, before the unit can be returned to
service. The susceptibility level does not designate or indicate the
ability of the component to force the unit from service for extended
periods; rather, it reflects only the components propensity to dam-
age and/or deterioration.

Owners will often identify potential problem areas in their units


at the first (warranty) inspection. At that time, the manufacturer will
make a detailed examination of the unit and be able to identify to
the operators any area or areas that have shown deterioration levels
above the norm for the units. The owner should then decide to
monitor these components, make a record of the existence of any
damage present, and define or record the extent in some manner.
The owner may also elect to carry replacement parts as inventory
spares in the event they are required.

Monitoring efforts, when undertaken, should be concentrated on


the most susceptible locations and components. A monitoring system

55
Turbine Steam Path Maintenance and RepairVolume One

can be established in conjunction with the unit manufacturer, who is


normally interested in such information for future designs.

Steam seals are one series of components of the steam path that
do not have a significant susceptibility to cause a forced outage
though their performance can influence the operation of the unit
considerably. High leakage rates in seals will cause extensive losses
to an owner in terms of higher fuel costs. It is generally prudent to
have replacement seals or gland rings available at each outage, and
the owner should know the limit at which deterioration (increased
leakage area) can no longer be accepted. The owner should be pre-
pared to replace worn seals whenever a unit is available.

Many seals are produced as an integral part of another compo-


nent. These cannot be easily or economically replaced. However,
refurbishment actions can often be taken to improve their effective-
ness without major replacement. Recent advances in repair and
refurbishment technology have added considerably to the ability of
owners to refurbish existing components. This is particularly true for
rotating blades (see chapter 8). Such refurbishment techniques can
often reduce maintenance costs significantly, and return the ele-
ments to a fully acceptable mechanical condition within periods of
time sufficient to support a normal maintenance outage.

While these refurbishment techniques are often able to return


steam path elements to an acceptable condition, it may be necessary
to plan that such refurbishment is undertaken during periods of low-
load demand (when unit output is not required or can be replaced
from other sources and the unit can be made available for this work).
If the refurbishment process is protracted, the resulting extended out-
age, and the cost of purchasing replacement energy may more than
offset any potential savings from refurbishment. Some refurbishment
techniques may be undertaken on-site and in place; others may
require the components be removed from the unit.

56
Consideration of a Turbine Steam Path Maintenance Strategy

INSTANTANEOUS
DAMAGE OR FAILURE
An instantaneous failure is one that occurs when the owner has
no prior or minimal warning from a previous outage or observations,
to the extent preventative action could be taken. In such cases, a sig-
nificant failure is suspected from a change in operating characteris-
tics and a unit is opened for inspectionand the failure is con-
firmed. At this point, the owner must make some determination
regarding the actions required to return the unit to an acceptable
condition, and then to service.

The instantaneous failure can be severe enough that refurbish-


ment actions required are self-evident. Other failures or damage
leave a degree of options available. In some instances the unit can be
returned to service for an acceptable period with partial repair under-
taken. A final decision concerning required corrective actions may be
influenced by diverse factors such as the amount of reserve generat-
ing capacity on the system and the cost and availability of replace-
ment or incremental power from other sources. The owner is also
influenced by the extent to which replacement parts are available.

The ability to refurbish the damage within an acceptable time


must be considered, as must the cost and potential of success likely
to be achieved. The ultimate selection of corrective action will most
probably involve at least the following factors:

Parts. Are replacement parts available from inventory? How long


would it take for such parts to be manufactured and installed? Can
the existing components be repaired or refurbished?

Power. Can the unit be operated with the failed elements


removed? What is the level of power lost, and what effect will this
have on unit heat rate and availability? Will loading limitations be
required to protect the remainder of the unit? If the unit is operated

57
Turbine Steam Path Maintenance and RepairVolume One

without the failed components, will the steam conditions to the unit
need to be reduced, or will the unit be load limited?

Time. Will such refurbishment be less expensive than replace-


ment, in terms of both initial cost, and the time to complete the
work? What are the possible effects on unit availability?

The effect of a major component failure or deterioration con-


cerns any owner. The owner will normally suffer a financial penalty
associated with such damage, through the cost of rectifying the situ-
ation and/or additional generating costs during the units unavail-
ability. It is therefore necessary, in such circumstances, to determine
the most cost effective method of correcting the situation and pre-
venting reoccurrences.

Often when new parts are installed after some form of failure,
possibly the removed parts can be refurbished and carried as inven-
tory spares. This would allow a similar situation in the future to be
dealt with in a more rational and expeditious manner. This is partic-
ularly so if the components are prone to failure, or the unit is one of
a multi-unit installation. The owner of a unit that has suffered some
form of damage or failure should always consider this approach.

It is not possible to formulate any rules to assist in decision-mak-


ing related to instantaneous or sudden deterioration. This damage is
only detectable from a change in the mode of operation of the unit,
or its presence is determined at maintenance inspections when the
steam path is exposed. At the time damage is found, the owners must
determine the most expedient course of action. Such a determina-
tion may be to replace the elements or to undertake refurbishment,
should this be possible. The logic for such an evaluation indicates
the analysis and decisions required to establish a repair procedure
(Fig. 1.11.1).

Often circumstances of instantaneous failure or damage mean


insufficient time to make extensive plans without extending the outage,

58
Consideration of a Turbine Steam Path Maintenance Strategy

INSTANTANEOUS DAMAGE,
UNIT FORCED FROM
SERVICE.

Have other portions of the unit


suffered consequential damage?
Can temporary Is refurbishment
repairs be an option?
Can the unit continue to operate undertaken?
if these components
are removed?
Are replacement Are replacement
parts required? parts available?
How critical is this unit
to system security?

What are the costs


involved in each
option?
CORRECTIVE ACTION
DECISION
What is the time
frame for completing
each option?

Fig. 1.11.1The evaluation of a failureFigure


and the1.11.1
considerations required to arrive at an
acceptable solution.
The evaluation of a failure and the considerations required
to arrive at an acceptable solution.

and the owner is forced into making a decision based on the most
expeditious course at that time. In the long term this may not be the
most suitable for rectifying the condition of the unit, and if doubt
exists, a decision and action plan should be delayed until a more favor-
able corrective action is possible.

FACTORS CONTRIBUTING TO
GRADUAL DETERIORATION
A gradual deterioration is one that occurs due to some mecha-
nism slowly lowering the performance potential of the unit. Such a
situation can be monitored from station instrumentation or is a con-

59
Turbine Steam Path Maintenance and RepairVolume One

dition measured and noted at planned outages. It is also a condition


that allows monitoring so acceptable corrective action can be deter-
mined and planned corrections made before the condition reaches a
level the unit is unable to operate in either a safe or efficient manner.

There are principal and various mechanisms that can contribute


to gradual deteriorationall possibly leading to failure:

the operating environment within the unit

stress levels

material properties

chemical contamination within the working fluid

carry-over from other parts of the system

Here a brief survey of these factors will be considered.

Operating environment considerations


The steam path is designed to operate and achieve its output
under stipulated steam conditions. The flow quantity required for a
specific outputthe initial, reheat, and exhaust steam conditions
dictate stage geometry, construction materials, and design details.
System economics tend to dictate unit and cycle configuration.

While steam temperatures and pressures at operational speed


may introduce high levels of stress on components under normal
operating conditions, modern design technology can help ensure
these will not induce failure if the components are correctly
designed. Under normal circumstances these are not factors that
would cause elements to deteriorate or fail.

There are, however, instances where transient conditions (partic-


ularly those of initial and reheat temperature) will induce high ther-

60
Consideration of a Turbine Steam Path Maintenance Strategy

mal stresses in the components, initiating failure. The components


most affected by these transient temperatures are the rotors, rotating
blades, and casings. Manufacturers normally provide guidelines to
operators defining a means of assessing potential damage accumu-
lation as a function of the rate of temperature change. A typical curve
of life expenditure indices for a high-pressure rotor is shown in
Figure 1.12.1.

SHADED AREA IS
BORE STRESS LIMIT
(NOT TO BE ENTERED
DURING INCREASES)

VALUES ON CURVES
ARE LIFE EXPENDITURE
IN PERCENT PER CYCLE

Fig. 1.12.1Life expenditure curve for a high temperature, high pressure rotor.

Initial steam pressure does not normally cause concern in terms


of placing the unit at risk. However, most manufacturers will place a
limit on the amount of increase for which their units are suited, and

61
Turbine Steam Path Maintenance and RepairVolume One

the number of hours of operation at these higher pressures. A pres-


sure limit is normally stated as a percentage of the initial design
value. Such a condition is generally of more concern in industrial
installations where other sources of steam may become available.
Under these conditions if applied without a full evaluation of the
unit (and possible steam path modification), it could induce failure
due to one of several mechanisms.

For condensing units, minimum pressure must be maintained in


the condenser. The last stage blades of large utility units are a vari-
able pressure ratio element (this ratio also varies with radial height).
If the exhaust pressure is raised too high (in excess of 6 Hg, or inch-
es of mercury), there is a possibility of inducing flutter type vibra-
tions within the L-0 blades. There is also a possibility of deterioration
caused by re-circulation through the lower vane sections of the dis-
charging steam. This could occur because of the reduction in volu-
metric flow. Frictional heating of these last stage blades will cause a
change in their operating temperature, and therefore could modify
their vibration characteristics, which are temperature sensitive.

The one environmental effect that causes some level of deteriora-


tion in most condensing units is moisture-impact erosion. However,
in a well-designed unit, this type of damage should not be an impe-
tus to change blades. In fact, few units suffer material loss and dete-
rioration to the extent blades cannot operate for 30 years. It is possi-
ble that towards the end of this period, there could be a significant
deterioration in stage efficiency, but this is not a consideration of
forced outage. At this time a unit normally operates at lower load fac-
tors and is not as important to the security of the power system.

More problems are induced by the methods of shield attach-


ment, which is a process involving the use of heat. It can expose the
blade material to the effects of heating cycles.

62
Consideration of a Turbine Steam Path Maintenance Strategy

Stress level considerations


Stresses are induced in the unit during normal operation due to
the steam environment, the effects of centrifugal loading, and the
bending loads induced by the expanding steam.

As force develops on the blade rows, it is transmitted through the


rotor to the generator or mechanical device that the turbine drives,
causing torque in the shaft. Stresses in the stationary portions are of
considerably less magnitude than those in the rotating components.
In general, stationary components do not fail due to high-level stress,
but there are instances in which stationary blades can suffer high
vibratory loads and fail. Also, parts affected by creep will occasion-
ally deflect axially to an unacceptable level.

The dangerous levels of stress induced in the rotating portions of


the unit are most often those caused by stress concentrationa func-
tion of the geometry of the components. During the design phase, it
is normal to evaluate the effect of stress, and modern design meth-
ods are fully capable of determining even the most complex forms.
Unfortunately, design conditions may not be maintained for long
periods. There are many regions of the rotating component where
stress concentration can be the result of mechanical damage, where
impact or other types of deformations are produced as the result of
operation. Such deformations should be looked for and some evalu-
ation made of their possible consequences. Non-destructive exami-
nation should be undertaken of such discontinuities when they are
found.

Pressure impulses generated in the steam path have the potential


to amplify the dynamic loading applied to the various components.
The design process attempts to recognize, predict, and quantify these
potential effects, minimizing their damage-causing potential. How-
ever, they continue to occur and units are often forced from service
as a consequence of their presence.

63
Turbine Steam Path Maintenance and RepairVolume One

Material property considerations


For each component comprising the steam turbine, materials are
selected and their properties evaluated for suitability. Materials are
selected to help ensure the unit will operate for its entire life with the
minimum probability of failure or need for replacement, if operated
to design specification. There are, however, some components con-
sidered consumable, and it is the intention of design that these will
be replaced as required by the user.

The designers of steam turbines evaluate materials and establish


the variation of long life properties at various required temperatures.
During the design phase the components are selected and dimen-
sioned to achieve an acceptable factor of safety at the predicted
stress levels and local environmental temperature.

To ensure that materials meet their design specified require-


ments, the manufacturer stipulates chemical composition, material
properties, method of manufacture, heat treatment, grain structure,
and level of non-metallic inclusionsall of which could affect life
and performance. When the materials are received from the suppli-
er, and before use, the manufacturer will evaluate them. The manu-
facturer should not knowingly use any that do not meet the stipulat-
ed requirements.

All high temperature components (above 900F) and many oper-


ating at high stress levels are particularly susceptible to variations of
material properties. These components in particular should be eval-
uated in detail and not produced from any material that does not
meet the highest standards as specified by design.

One aspect of material suitability that should be considered is


many material properties, especially at high temperatures, are nor-
mally determined from short time test data. If the material is of a new
design, there could be some small margin of error in the stated val-
ues. A manufacturer will normally allow for this possibility as a fac-
tor of safety in the design.

64
Consideration of a Turbine Steam Path Maintenance Strategy

A designers additional concern for components designated for


high temperature operation is that material properties in these envi-
ronmentsoperating at high stress levelscan deteriorate with
time. This is an acknowledged fact and many components must have
either an adequate factor of safety to prevent significant deterioration
(blades), or be classified as a consumable (studs and nuts).

Considerations of chemical contamination


There are two aspects of chemical contamination that need to be
considered:

Possible deposits of compounds on steam path components


that will lower the efficiency of expansion and therefore
introduce extra aerodynamic losses within the unit. This is
undesirable in terms of operating cost, but is not an effect
that will lead to mechanical deterioration. It should be a part
of each normal maintenance outage to remove these
deposits by some suitable cleaning procedure

Contaminants could gain access within the total thermal


cycle and form corrosive compounds. These compounds can
be carried over from the boiler and other portions of the
cycle, and will eventually be deposited on some steam path
components. Once deposited in the steam path, they will, at
some locations, be the constituents of failure in the unitan
aggressive chemical compound, tensile stress, and an elevat-
ed temperature. These contaminants can react with con-
struction materials and could eventually lead to mechanical
failure, forcing the unit from service

65
Turbine Steam Path Maintenance and RepairVolume One

Considerations of carry-over
from other portions of the system
There are two types of carry-over into the turbine that need to be
considered:

The possibility of metallic or other solid debris carry-over


from other parts of the system. This is concerned principally
with the flow of solid particles, such as a weld bead from the
boiler. This is normally most severe during initial start up,
commissioning, or during early stages of operation. During
these periods it is common to use a fine mesh screen over the
main and intermediate valves. Also, after some years of oper-
ation, hard oxide scale exfoliated from boiler tubes will enter
the unit and cause solid-particle erosion (Figure 4.3.3 in
chapter 4 shows a fine mesh screen with contained weld
bead)

There is also the possibility at any time during operation for


water to collect in a pipe above a choked drain. Under certain
circumstances, this water can be carried over into the unit. The
extent to which this will cause damage depends upon the
quantity of water in the line, the manner of its entry to the
steam path, and the extent and duration of the ingestion

MONITORING DAMAGE
AND DETERIORATION
There are certain types of damage or deterioration that, when
they occur, may not force a unit from service immediately but could
have a long-term effect on availability. These are recognized as grad-
ual deteriorating mechanisms. When such gradual deterioration is

66
Consideration of a Turbine Steam Path Maintenance Strategy

detected, it is often necessary and expedient to return the unit to


service. This represents a situation under which the operator can,
and often should, initiate some form of monitoring program
designed to establish the deterioration rate. If necessary, corrective
action can be planned for a future outage. There are some forms of
deterioration that can be accepted indefinitely, but this is usually
done at some loss in unit efficiency.

One of the most complex and potentially controversial decisions


the maintenance staff will be required to make when responsible for
steam turbine operation and performance is: At what stage in their
operating life has the condition of the components of the turbine
steam path deteriorated to the extent corrective action must be
takeneither to restore the efficiency to an acceptable level or to
prevent failure which would force the unit from service?

The factors that must be considered when damage is found, and


the decisionspossibly backed by financial evaluationsinclude
the following:

Does the situation require immediate action to correct this


deterioration, or can the unit continue to operate at accept-
able levels of performance (efficiency and reliability) with
the damage it has sustained?

If corrective action is considered necessary, is it appropriate


to replace the components with new, or can existing ele-
ments be refurbished? If refurbishment is considered ade-
quate, can the work be undertaken on-site and in place, or
must the components be removed from the unit?

Since the mid 60s, vibration signature analysis has provided an


indication of the overall condition of a unit, and regular monitoring
has allowed owners to detect and possibly identify pending prob-
lems. However, signature analysis provides evidence of change in
operating characteristics once they have deteriorated to a point

67
Turbine Steam Path Maintenance and RepairVolume One

where possibly significant damage has occurred. Decisions related


to the timing of corrective action and outages are made easier if the
condition of specific components, with a known potential to deteri-
orate, are monitored at regular inspections and their deterioration
rate established.

A suitable method of monitoring, and one that can be conduct-


ed by plant operators without the use of any expensive equipment,
is to examine the unit when it becomes available at regular mainte-
nance inspections. At that time, the unit should be examined for evi-
dence of abnormal or excessive deterioration.

The condition of the components should be quantified and


recorded through measurement, castings, and photographs. It is
unfortunate, but with few exceptions it is not possible to make an
examination at other than periodic inspection outages. This means
the steam path may only be available for detailed examination on a
normal outage cycle. Because of this, it becomes necessary to exam-
ine components in detail at these outages, note the extent of all dam-
age or deterioration, refurbish to the extent possible, and then plan
to have replacement parts available for installation at subsequent
outages.

Any component of the steam path that deteriorates can be sub-


ject to two levels (or rates) of deterioration. First is that which occurs
suddenly, possibly instantaneously, as the result of some transient
condition or other phenomena of operation. These phenomena can
include the failure of an upstream component that causes conse-
quential damage, or the ingestion of large quantities of water. Such
damage is normally severe and may give little or no warning of its
pending occurrence. Therefore, little or no indication is offered that
the unit should be shut down prior to the damage. After the damage
has occurred (and depending upon its severity), there may be a small
increase in the vibration level or minor change in other operating
characteristics to provide evidence of its occurrence. Therefore, the

68
Consideration of a Turbine Steam Path Maintenance Strategy

damage may not be detected for some period, during which time the
condition could deteriorate further.

The second level of deterioration is that which occurs as a grad-


ual effect, such as the growth of a crack or component material loss
due to some form of erosion. This second level of deteriorationthe
gradual worsening of the component conditionwill often allow the
owner to select and plan his remedial actions, and to make replace-
ment parts and services available to meet an outage schedule. Such
a refurbishment outage may require, however, that a normal mainte-
nance period be extended to allow satisfactory reconditioning.

The following sections consider various causes of gradual deteri-


oration and, where possible, provide some guidelines that will assist
operators in determining the extent to which components need to be
replaced or refurbished. Unfortunately, there are no general rules
that apply to any set of damage conditions. Obviously, the most sat-
isfactory method is to replace or refurbish components when dam-
age is first noted. However, this is not only expensive in terms of the
replacement cost, but could require an extended waiting period for
such parts to be delivered or reconditioned. Normally, a plants out-
age plans do not support such a delayed decision. There are
instances when damage can be accepted and monitored but this
possibility must be evaluated in each case.

The mechanism of gradual deterioration


Those principal and various mechanisms that can contribute to
gradual deterioration (possibly leading to failure) are considered in
the previous section Factors Contributing to Gradual Deterioration.
When damage due to these forms of deterioration is found, it is rec-
ommended that the condition be monitored so corrective action can
be taken before the unit is forced from service.

69
Turbine Steam Path Maintenance and RepairVolume One

Monitoring techniques for gradual deterioration


There are certain types of damage or deterioration which, when
they occur, may not force a unit from service immediatelybut
could have a long-term effect on unit performance. When gradual
deterioration is initially detected, it is normally necessary and expe-
dient to return the unit to service. This represents a situation by
which the owner can (and often should) initiate some form of mon-
itoring program to determine the deterioration rate. If necessary, cor-
rective action can be planned for some future outage. Some forms of
deterioration can be accepted indefinitely, but this offers the possi-
bility of a loss in unit efficiency.

This section uses examples to discuss typical monitoring tech-


niques that can be applied to portions of the steam path compo-
nents. They provide guidance of methods suitable for operators to
use in establishing the rate of deterioration within the unit. They also
provide, where possible, some guidance when conditions are dete-
riorating to an unacceptable level and corrective action should be
considered.

In providing this guidance it must be remembered each unit is


unique and represents a different set of parameters. Judgment is very
much dependent upon the mode in which a unit is operated. Also,
the owners previous experience is often a good indication as to the
danger posed by any situation. With these provisos, then, the follow-
ing observations and situations should be considered as a guide, sub-
ject to calibration from system, unit, and manufacturer experience.

Moisture-impact erosion. In general, moisture-impact erosion


will not by itself cause sufficient material loss that a blade will need
to be replaced. Damage and material loss that occur in the outer
portions of the blade vane should not impact upon element reliabil-
ity. This is true to the extent the erosion occurs on the inlet edge suc-
tion face.

70
Consideration of a Turbine Steam Path Maintenance Strategy

There are, however, certain aspects of erosive material loss that


can cause concern, and possibly require some form of corrective
action:

Excessive erosion below the erosion protective shield

Erosive undercutting of tenons attaching a coverband

Severe erosion beyond, or behind, the erosion shield

Severe erosion endangering the attachment of the erosion


shield

Severe erosion endangering the attachment of a vibration-


tuning device

Moisture-impact erosion is a time dependent phenomena,


though the damage or material loss does not vary linearly with time.
The curve (Fig. 1.13.1) indicates erosion penetration () as a func-
tion of time in hours and shows the approximate time in years. The
unit load factor is 70%. This curve can be used to predict future pen-
etration as a function of time as the value of changes. This will

3.0
Erosion Penetration

2.0

1.0

0
0 5 10 15 20 25 30
1 Year Rate. Years

Fig. 1.13.1The erosion penetration,Figure


as a function
1.13.1of time.
The erosion penetration, as a function of time.

71
Turbine Steam Path Maintenance and RepairVolume One

allow some estimation of further damage that could occur. (This rela-
tionship between penetration and time is considered in greater
detail in chapter 3.)

Localized or secondary erosion can be examined and its possi-


ble effects estimated from the dimensional changes this produces on
the blade. Blades subject to moisture-impact erosion are also ele-
ments that are tuned. These elements normally have high stress lev-
els and their frequencies can be sensitive to changes in blade sec-
tion. Figure 1.13.2 shows a blade row that has suffered some small
level of local secondary erosion that removed a greater amount of
material at one specific location.

Fig. 1.13.2Exhaust blades showing the variation of erosion


penetration in the outer flow sections.

72
Consideration of a Turbine Steam Path Maintenance Strategy

An example of monitoring the last-stage blades of a unit with pre-


dicted heavy erosion is shown in Figure 1.13.3. Here the mean val-
ues of erosion in the outer section is shown for the turbine end A and
the generator end B. In addition, the worst blade is shown. These
blades were eventually replaced/refurbished after 217,000 hours of
operation as part of a general refurbishment program. The condition
of the blades was monitored using casts and measurements.

Fig. 1.13.3Erosion penetration history of a turbine back end with heavy erosion.

Trailing edge effects of erosion. Erosion of the exhaust stage


blade trailing edge occurs due to moisture particle re-circulation
through the root section. This material loss occurs when cooling
sprays are used to cool the exhaust hood. Figure 1.13.4 shows a typ-
ical cross section of an exhaust blade near the discharge edge of the
root profile. The damage from this type of erosion removes material
at a location where vane stress levels are high. Therefore, only rela-
tively small levels of material loss can be tolerated.

No general rules exist for estimating failure potential. To determine


such detail, an in-depth stress analysis (probably by finite element

73
Turbine Steam Path Maintenance and RepairVolume One

methods) is required. However, the operator should make a true eval-


uation of the stress levels and concentration potential once the pene-
tration d has extended to 0.050" or half the vane thickness t at dis-
chargewhichever is less. Such an evaluation should be made prior to
reaching a dangerous level of penetration. If analysis proves the mate-
rial loss to be dangerous, efforts should be made to limit operating con-
ditions and cooling water flow quantities to the extent the damage is
limited. Such blades will eventually require replacement.

Fig. 1.13.4Erosion of the discharge edge with


erosion d extending to a limiting value d2. The
water discharges from the groove in a direction
w.

When such damage is initially noted, it can be monitored by


means of a plot of penetration d as a function of hours of operation
with the cooling sprays operating. If the material loss becomes
severe, the owner should consider plans to replace the blades.

Fortunately, the last stage blades (and particularly the discharge


edge) can be inspected easily. When this type of damage reaches a
critical level, such examination should be undertaken regularly,
including the application of non-destructive methods.

Solid-particle erosion. When oxide scaleparticularly hard and


abrasiveis carried into the turbine steam path from the steam gen-

74
Consideration of a Turbine Steam Path Maintenance Strategy

erator or reheater, it has the potential to remove material and affect


both the mechanical integrity and thermodynamic efficiency of the
steam path. Such scale can remove material from the stationary and
rotating blades and, after being centrifuged to the casing, can rebound
and remove material from the tenons attaching the coverband.

While the stationary blades may suffer little adverse influence


from stress effects, they have the potential to significantly influence
the efficiency and mechanical integrity of the remainder of the
expansion. The following rationale applies to material loss from sta-
tionary blades:

The quantity of steam flowing through any stage, and the dis-
charge pressure from a blade row, is a function of the dis-
charge area from that row. When the affected stage is of a
two-flow arrangement in any pressure section, there could
be an unequal sharing of the flow between the two halves.
This occurs when erosion is not equal on the two halves of
the flow

The effective discharge angle from any throat formed by a


blade pair is a function of the ratio throat opening to
pitch. Figure 1.13.5 illustrates the variation in discharge
angle (or ratio) O/P as material is removed from the dis-
charge edge due to solid-particle erosion

Wear is normally not uniform along the radial height of the vane.
The effects of uneven wear will be to cause a significant variation of
discharge angle on the stationary blade and the magnitude of the
steam impulse from it.

While the rotating blades are subject to the same considerations


as apply to the stationary blades (in terms of discharge area and
angle), they are also subject to centrifugal loading. This means con-
siderations of stress are often more significant. Fortunately, the scale
that causes erosion (because it experiences the centrifugal actions

75
Turbine Steam Path Maintenance and RepairVolume One

imposed by the rotational effect) will tend to attack the outer regions
of the blade, where centrifugal stresses are lower. However, if the
affected blade elements carry a coverband, the bending stresses
imposed on the outer portion of the blade vane can be severe. It is
possible for these stages to suffer damage.

Fig. 1.13.5The variation in the ratio O/P as material is lost from the
discharge edge due to SPE.

An insidious aspect of solid-particle erosion (SPE) is the erosion


of tenons between the vane tip and the underside of the coverband.
This erosion can weaken the coverband attachment and possibly
lead to coverband detachment.

SPE is time dependenteffects are dependent upon operating


hours. However, units with a large number of start ups are more sus-
ceptible to this type of damage because that is when scale is exfoli-
ated from the boiler tubes. A suggested method of monitoring pene-
tration d is shown in Figure 1.13.6. It shows that monitoring a sim-
ple calculation of the effects on stress levels in both the blade vane
and shroud band will provide some indication of when corrective
action needs to be taken, usually in the form of new blades.

76
Consideration of a Turbine Steam Path Maintenance Strategy

Depth of erosion
penetration 'd'
d1
d2
d3
d4

Figurethe
Fig. 1.13.6Showing 1.13.6
measurement of
SPE penetration
Showing at the vane inlet
the measurement edge.
of SPE
penetration at the vane inlet edge.

Crack indications in the blade vane. The appearance of a crack


in the rotating blade vane or transition from the vane to root platform
is sufficient warning that the blades may have reached the end of
their useful life. Fracture mechanics methods exist and can be
applied to indicate the rate of crack growth and eventual fracture.
However, it is assumed that any crack has sufficient potential for cat-
astrophic failure in a steam turbine and the element should not be
returned to service.

One necessary exception to this was the case of last stage blade
elements in which cracks appeared at the tie wire hole. They were
evident below the braze material. These cracks were initially found in
three elements. This was a utility unit, whose continued operation
was essential to meet demands for power and heating steam. After the
utility ordered replacement blades, it returned the unit to service and
conducted monthly inspections and measurements of the cracked
elements. These condition reviews were at weekend outages, when
the unit could be made available for inspection from within the

77
Turbine Steam Path Maintenance and RepairVolume One

exhaust hood. After some months, a weld repair of the cracks was
made on-site and in place. The blades continued to operate without
problems. A year after the weld repair, the blades were replaced. At
that time the cracks had not reappeared, but the utility elected to
replace the blade row with new elements. The history of these blades
from initial discovery of the crack is shown in Figure 1.13.7.

On many units, a crack appears at the fillet radius between the


vane and root platform. In general, this type of crack cannot be
accepted. However, there have been instances in which such cracks
have been removed by grinding and/or polishing and the unit
returned to service. The extent to which this is possible is a function
of the depth of the crack. Such a practice is not recommended as a
permanent fix. Instead, the blades should be removed and
replaced, as the risk level is high.

Extent of
0.5 Braze In situ weld repair
L
at week 16
Measured length of crack.

0.4 B
Tie wire A
hole

0.3 Possible crack propogation


curves "A" and "B".

0.2
4 8 12 16
Weeks of operation.
Initial finding of crack,
and return to service

Fig. 1.13.7Measured crack growth


Figure 1.13.7in a last stage blade. The
crack initiatingcrack
Measured at the tie wire
growth in ahole. See figure
last stage blade.1.8.17.
The crack
initiating at the tie wire hole. See figure 1.9.17.

78
Consideration of a Turbine Steam Path Maintenance Strategy

Diaphragm vane cracks. Diaphragm vanes can crack because of


a variety of phenomena that do not normally represent a serious sit-
uation. However, a crack initiating at the discharge edge can grow.
The edge can become unsupported and sufficient flutter can be
introduced to cause the discharge to break away. Therefore, such
cracks should be repaired. Depending upon the selected weld filler
material, such a repair can be undertaken on-site and in place using
a weld rod selected for the temperature of the stage. Low-pressure
unitsparticularly those manufactured by welding techniqueswill
often develop cracks running across the vane. The growth of such
cracks should be monitored and the element repaired at some suit-
able outage. Because it is relatively easy to make such a repair, it is
not necessary to monitor these cracks. Under normal circumstances,
the diaphragms should be repaired and returned to service.

Seal system wear. Seals within the steam path provide a con-
striction between the stationary and rotating parts. Steam that leaks
(or expands) past these constrictions represents wasteful expansion
and a reduction in the energy generated. Seal wear should be mon-
itored, and as such wear becomes excessive, seals should be
replaced when and where possible.

The cost that excess clearance represents can be calculated in


terms of geometry of the stage, thermodynamic conditions at the seal
point, and fuel cost. Owners must understand the cost of incremen-
tal power as a function of seal clearance, and be prepared to
upgrade these seals as wear increases. It is normal practice within
the steam turbine industry to monitor clearances throughout the unit
each time it becomes available for such measurements. This infor-
mation is valuable in a variety of ways. One of the most important is
expensive fuel, which can be a predictor of when seals should be
changed. Losses associated with seal wear are considered in chap-
ter 10.

Corrosive damage. Corrosion is the result of chemical action in


the steam path caused by corrodents introduced into the cycle by

79
Turbine Steam Path Maintenance and RepairVolume One

external sources. Such corrosion can occur on visible surfaces dur-


ing normal inspection outages. Surface-type damage is not normally
significant in terms of the reliability of the unit. The more serious
aspect of corrosion occurs at points within the steam path where
hideouts occur, such as at the shrink fit between the wheels and
shaft, and the tie wires and tenons that attach the coverband. Units
should be inspected (as far as practical) for this type of damage
when damage is discovered, components should be replaced.
Unfortunately, much of this type of damage is usually invisible and
can neither be seen nor adequately monitored.

Surface corrosion is less severe in terms of unit performance, but


can indicate hidden corrosion. Therefore, when surface damage is
found, it is a good practice to use photographs to monitor its extent
to compare deterioration and consider future damage at hideouts.
Unfortunately, the corrodents causing surface damage may not be
the same as those that concentrate in hideouts and lead to compo-
nent cracking.

Checking for crack initiation in hideout regions is complex and


often expensive. Therefore, when such damage is suspected, an eval-
uation should be undertaken. If visual inspection confirms the possi-
bility, an ultrasonic examination can be undertaken. The shrink fits
between wheels, and central spindles are particularly difficult to
access and inspectit can represent one of the more complex eval-
uations expected of the operator engineer.

Fretting damage. Relative movement between components caus-


es fretting corrosion as pressure develops between their contacting
surfaces. By definition these surfaces are rarely visible and the dam-
age is hard to detect. However, when such damage is found, its
extent should be monitored. It has the potential to loosen fits, which
if design does not allow, can cause a change in the mechanical
dimensions of the component and aggravate what can be an already
serious situation.

80
Consideration of a Turbine Steam Path Maintenance Strategy

There are situations in which relative movement between sur-


faces is required. This means that conditions can occur in which the
oxide products of fretting can cause a designed sliding surface to
become solid and defeat the design requirement.

When this type of damage occurs, it is best recorded by means


of both photographs and casts of the surface, if accessible, from
which depth and affected area can be determined in extreme cases.

Water washing erosion. High-pressure water (or water at high


velocity) has the potential to cause significant damage within the
steam path. It can also compromise the integrity of the main struc-
tural components of the unit.

Washing or wire drawing type damage on a seal surface can


reduce the stage efficiency. Also, the leakage flow can interfere with
the free flow of the expanding steam, possibly causing circulation
losses to be introduced into the flow path. When this type of dam-
age is discovered, it is normal to assess the ability of the component
to continue to form effective seals, and if necessary, make repairs.

When washing damage is not sufficient to require immediate


repair action, a monitoring program (as a function of time) should be
introduced. This is a type of damage best recorded by photographs.

The steam path requires exact dimensions of the passage through


which the steam is expanding. Washing damage that modifies the
shape of the flow passage must be monitored, then repaired or
replacements parts used as soon as the steam path integrity is com-
promised.

The other type of material loss mechanism that requires some


monitoring (particularly in older units) is the combined
washing/impact material loss (Fig. 1.13.8). Here, material is
removed from a structural component to the extent wall thickness is
reduced and the vessel possibly weakened. Damage of this type is
not common, but when it does occur, the results can be severe.

81
Turbine Steam Path Maintenance and RepairVolume One

Fig. 1.13.8Combined washing and impact erosion on the cast outer sidewall of a
diaphragm.

REPLACEMENT PARTS
STRATEGY AND SUPPLY
Alternate suppliers of steam path components have evolved, and
have had a dramatic effect on the manner in which many owners
plan and undertake their maintenance programs. The most signifi-
cant of these changes include:

the owners ability to carry a smaller inventory of consum-


able parts. This minimizes the cost of maintenance by reduc-
ing the funds tied up in components not required for years

the ability to order known, required parts weeks, rather than


months, before an outage. This reduces costs and concerns
regarding on-time delivery

82
Consideration of a Turbine Steam Path Maintenance Strategy

the owners ability to make replace-or-repair decisions for


unanticipated damage and deterioration when the unit is
opened and its condition determined at the outage

Despite these advantages, the owner will not be able to order all
required material for each outage and every contingency after the
unit is opened. However, many of the emergency parts previously
unavailable from inventoryparts not generally carried as stock
can be made available from alternate suppliers when inspections
establish their need.

83
Turbine Steam Path Maintenance and RepairVolume One

REFERENCES
1. Dewey, R.P., and N.F. Rieger. Survey of Steam Turbine Blade
Failures, Research Project TIE, Final Report, March 1985

2. Open letter from General Electric: Extending Re-inspection


Intervals-GE Turbine Rotors, Made available at the EPRI
Steam Turbine-Generator Workshop, Milwaukee, Wisconsin,
August 1995

3. Sanders, W.P. The Review, Assessment, Refurbishment and


Efficiency Upgrading of Steam Turbine Units, Canadian
Electrical Association, Halifax, Nova Scotia, 1985

4. ABB Power Generation Ltd, Baden Switzerland: Overhaul


Plans for Large Steam Turbines-HTGD 690 218 E and HTGD
90 700E

5. Recommendations for the Inspection of Steam Turbines: -


VGB - R115 Me. VGB Essen, Germany

6. Sanders, W.P. Potential for Manufacturing Problems to Cause


Steam Turbine Blade System Failure, Turbomachinery
International, Vol. 27, No 7, September 1986

7. Sanders, W.P., and W.R. Southall. A Guide to Mechanical


Condition Assessment in the Turbine Steam Path, EPRI Steam
and Combustion Turbine Blading Conference, Orlando,
Florida, January 1992

8. Timo, D.P. Design Philosophy and Thermal Stress


Considerations of Large Fossil Steam Turbines

9. Rieger, N.F. Blade and Rotor Reliability Improvement,


STI/ERPI Seminar, Rochester, New York, June 1991

10. Greco, S. Private correspondence

84
Chapter

2
Steam Path Component
Alignment and Stage
Spatial Requirements

INTRODUCTION
In order to achieve an acceptable level of performance from the
steam turbine, two things are essential:

individual components comprising the steam path are manu-


factured in accordance with the design specified requirements

components are arranged or assembled within the unit so


their spatial relationship, relative to the other components
with which they will interact, are correct

This compliance with design requirements will optimize the


energy conversion process and help ensure the structural reliability
of the system is at an optimum level. The spatial requirement within
the steam path will be considered in this chapter, along with factors

85
Turbine Steam Path Maintenance and RepairVolume One

influencing the relative positions of the various components one to


the other.

Note: For the purpose of this work the steam path is considered
to comprise the stationary and rotating blade rows, their means of
attachment, the rotors, and the casings. To complete the arrangement
of the components and help ensure a satisfactory design, it is neces-
sary to consider the arrangement made to minimize internal leakage
and the leakage occurring from within the casing to other portions of
the steam power cycle. In the case of low-pressure sections, leakages
into the system must also be considered.

Before considering details and requirements of the individual


components comprising the steam path, we must review their posi-
tional needs relative to each other, and those factors influencing
them when the unit is in operation. Unfortunately, the steam path is
available for measurement, adjustment, and gauging only in its cold
stationary positionconditions under which components must be
erected and aligned. When the upper half casing is assembled to the
lower and steam is admitted to the unit, the component parts change
relative to one other. These changes are due to various factors influ-
enced by steam temperature and pressure and by the mass and rota-
tional effects of the rotor system.

During initial manufacture (or repair or replacement of various


path components) certain spatial relationships in the axial, tangen-
tial, and radial directions must be met. Other relationships, by pref-
erence, should be met to assist in optimizing performance. The tur-
bine supplier will manufacture and assemble the component parts of
the steam path to ensure design requirements are achieved within
specified tolerance bands. The builder monitors this process. In addi-
tion, within the organizations of some purchasers, groups have a
responsibility for monitoring these operations.

It is not unreasonable to assume a unit, as originally supplied


and erected within the operators plant, will conform to design

86
Steam Path Component Alignment and Stage Spatial Requirements

requirements. Discrepancies with design specification must be


recorded and information supplied to the owner, so corrective action
can be taken as the opportunity arises.

A unit removed from service for inspection and maintenance


often requires some level of remedial work. Remediation aims to
return the components and their overall arrangement to conditions
as close to original as possible, consistent with preserving the units
performance potential. It is therefore valuable for plant maintenance
staff to have sufficient knowledge of the component parts and their
assembled requirements to determine the most appropriate course of
action in any situation where adjustment or corrective action is
required.

In terms of manufacturing requirements, the turbine designer


must provide exact definitions. These must be made available to the
manufacturing portion of the supplier organization to ensure design
requirements are first understood, and then met within specified tol-
erances. This information must be provided in such a form that it
allows design needs to be understood and followed by the manu-
facturing department without ambiguity. Design requirements will
also identify boundaries or engineering tolerances within which the
design requirements must be met. Tolerances specified by the
designer should reflect the limits within which components can be
manufactured and still achieve a level of performance consistent
with design predicted values. These tolerances must also reflect
achievable values for the processes and component being consid-
ered.

The operating engineers and technicians within the purchasers


plant normally do not have access to detailed design information
this is normally considered proprietary. They should, however, have
sufficient opportunity during the manufacturing and erection phase
to ensure that components meet design approval. They will normal-
ly have sufficient information on clearances, unit alignment, and the

87
Turbine Steam Path Maintenance and RepairVolume One

initial design conditions, and can reestablish conditions to ensure


correct spacing between the parts is maintained (at least in axial and
radial directions).

Engineering tolerances are selected to ensure that the design


expectations of performance levels are met, and where appropriate,
components can be disassembled for repair or replacement.
Tolerances should also help ensure components can be interchanged
within similar units in and between stations. Unfortunately, there are
times when the requirements of design are not sufficiently explicit,
or the unit operator has been provided insufficient information. In
such cases, maintenance staff is unable to evaluate all possible
courses of action and achieve the most appropriate corrective solu-
tions. When mechanical deterioration is found at maintenance out-
ages, it is necessary to evaluate the situation and then select correc-
tive actions that will restore the unit to a condition enabling it to be
returned to service and operate in a manner acceptable to the owner.
This condition may not represent a final repair condition, but will be
sufficient to make the unit safe to the extent it is able to generate
power at an acceptable level of output and efficiency.

The utility engineer cannot be responsible for ensuring the ade-


quacy of the design definition provided within the manufacturers
plant. However, the purchaser has an implied responsibility to
ensure this definition exists and that it is applied without compro-
mise during manufacture. Such definition should provide sufficient
detail that critical components and specific areas of risk to individ-
ual elements are addressed within the utilitys experience. It is also
realistic to expect that sufficient information is made available to the
owner allowing minor repairs and adjustments to be made on-site.
The operating engineer should be able to question design detail.

When individual unit components have been manufactured and


assembled, they can be erected to form the complete machine. For
smaller units, such assembly may take place in the manufacturers
plant and the components shipped pre-assembled, or the unit bro-

88
Steam Path Component Alignment and Stage Spatial Requirements

ken down to major components for shipment. With larger multi-sec-


tion designs, shop assembly is normally impractical (except in the
case of prototype units) and complete assembly would add consid-
erably to the total costs of the unit.

The process of site erection should result in a unit that is aligned


to maximize the units ability to generate power at an acceptable
level of efficiencyimmediatelyand to continue to operate reli-
ably over the designed plant life. Irrespective of the adequacy and
quality of design, if a unit is not installed correctly, it will introduce
operational problems reducing the overall utility of the unit to the
owner.

It is also necessary to interface the turbine generator with other


major pieces of equipment comprising the thermal cycle. These
important interfaces must be made so that they do not prejudice the
operation of either the turbine generator or other equipment.

Control of various areasand relationships within themis a


major characteristic in establishing the quality of the turbine steam
path. These areas are associated with the expansion and flow direc-
tion of working fluid throughout the steam path. Some of this area
control is achieved from a correct alignment of the steam path com-
ponents as they can affect area relationships to a degree. Also, while
alignment may not control the effective area in all cases, it can have
a considerable influence on the efficiency of energy conversion by
ensuring the steam is directed from blade row to blade row and
shock and incidence losses are minimized.

Areas and area control requirements are considered in some


detail later in this chapter. It is necessary to consider those phenom-
ena that influence alignment and the cold setting condition of the
componentsphenomena that are predictable and determined by
the design function for each unit. There are also phenomena that
affect alignment, are unpredictable, and occur normally as a conse-
quence of operating conditions. These have the potential to modify

89
Turbine Steam Path Maintenance and RepairVolume One

both material properties and physical characteristics of the individ-


ual components of the steam path, and ultimately affect alignment.

Stage terminology
A number of terms and phrases describe and define steam path
components. Problems in discussing blades are compounded by
manufacturers using terms not consistent one from another. In this
chapter (and this book), the terms used are defined in Figure 2.1.1,
and include the following:

Rotating bladethe total component, comprising the vane,


the root, and all other component attachments machined as
part of the total.

Blade vanethe airfoil portion, extending from the root block


to the tip. For rotating blades, this tip is the largest diameter
of the airfoil; for stationary elements, it is the smallest.

Profilethe form (or cross section) of the airfoil. This form


may be constant along the vane radial height (cylindrical) or
can vary from section to section (vortex) to accommodate the
changing steam parameters.

Root or fasteningthat portion below the vane and the load


transfer point from the vane to the rotor.

Root platformthat portion of the root block between the vane


and the root ligaments that carries the load-bearing surfaces.

Tenonsan integral part of the blade located at the tip diam-


eter. They are used to attach an inner or outer coverband.

Tie wire holethose stages that have a continuous tie wire


require an access hole in the vane. This is considered a por-
tion or characteristic of the blade.

90
Steam Path Component Alignment and Stage Spatial Requirements

Stationary bladesthe stationary blade elements mounted


directly into the casing or blade carrier.

Diaphragmcomprises three major elements that locate a


stationary blade row in the casing (inner or outer) and are
normally a welded or cast assembly.

Outer ringlocates the stationary blade elements in the cas-


ing, providing axial position and a steam seal from the high
to low-pressure side of the diaphragm.

Stationary vanesperform a similar function to the rotating


vanes and have the same stringent requirements for the profile.

Inner webthe material forming the expansion passage


inner surface and a steam barrier minimizing steam leakage
from the high- to the low-pressure side of the stationary
blade vane.

Profile
Tenons

Tie Wire
Hole
Vane
Vane

Root
Platform
Blade Root
Root Block

Root
Ligaments

Fig. 2.1.1Definitions of the2.1.1


Figure blade components.
Definitions of the blade components.

91
Turbine Steam Path Maintenance and RepairVolume One

The turbine unit, as installed, should present to the operator/


owner the optimum setting of the components, so when the unit
goes into service it will provide a maximum return on the money
invested in making the purchase and justify the selection among
competing bids. While many design-selected options of components
and arrangements within the steam path are a compromise between
competing requirements, produced within design specified toler-
ances, the final condition should be one allowing the unit to operate
and achieve acceptable levels of performance. Over time, this con-
dition will modify, efficiency will deteriorate, and the possibility of
mechanical damage and deterioration will arise, possibly affecting
unit availability.

Such deterioration means owners should establish a clear mainte-


nance strategy (explained in chapter 1). This allows them first to iden-
tify, and then correct, any nonconforming conditions as expeditiously
as possible. To do this the maintenance staff should be aware of:

the signs and indications of deterioration, and how to make


objective judgments of when corrective action is required

the potential or total deterioration that can occur if corrective


action is not taken

sufficient knowledge of the repair/refurbishment options


available, their possible cost, and the consequences of not
taking such action

Steam turbine performance depends upon some very demanding


thermodynamic relationships that are developed and maintained
within the expanding steam. Because relatively small changes in
component geometry or alignment significantly impact the total per-
formance of the unit, it is necessary that the critical characteristics of
these requirements are achieved and maintained within the individ-
ual elements. What is often not so clear to the operating engineer is

92
Steam Path Component Alignment and Stage Spatial Requirements

what is critical, and what can be ignored to a degree and accept-


ed before it becomes critical and requires corrective action.

Spatial requirements exist for steam path components in the radi-


al, axial, and tangential directions. Requirements of radial and axial
clearances should be observed to help ensure that the performance
of the unit is maintained at an acceptable level. This chapter dis-
cusses these and provides some guidance of these different require-
ments and the possible consequences of their not being achieved.

PREDICTABLE FACTORS
AFFECTING DESIGN CLEARANCE
Four basic phenomena (or characteristics of operation) are pre-
dictable, and influence the radial and axial clearances throughout
the steam path. These phenomena will affect the total alignment
within the stages, and at any location where stationary and rotating
surfaces are close. The extent to which these phenomena influence
the steam path design can be established (or predicted) by calcula-
tion and taken into consideration during the design phase when
selecting and arranging cold stationary clearances and bearing ele-
vations for the unit at initial assembly and subsequent alignment.

While the designer can calculate the hot running relationships


between the stationary and rotating parts, he or she can only mean-
ingfully define to the manufacturing and installation departments the
spatial relationships, clearances, and setting requirements as they
can be measured and set in the cold stationary condition. Therefore,
it is necessary first to consider the types of modifying effects that
need be evaluated, and then the degree to which they will influence
the cold stationary arrangement of the unit. There are four pre-
dictable (and always present) phenomena, whose evaluation

93
Turbine Steam Path Maintenance and RepairVolume One

remains part of the mechanical design, and therefore always consid-


ered in determining the initial settings.

The influence of these four phenomena must be predicted by


design, and their total effects compared on some form of clearance
chart. These values must then be used to select the optimum cold set-
ting at the various locations throughout the steam path with the final
cold settings selected to help ensure rubs do not occur, or reduce
the possibility of their occurrence during normal predictable opera-
tion. These four phenomena or areas requiring consideration
include:

the extent and shape to which the rotor will deflect vertical-
ly between the bearings due to its own weight, the extent this
is influenced by the temperature of the rotors, and the effect
of temperature on material properties

the differential axial movement that occurs between the


rotating and stationary portions of the unit due to thermal
expansion of these stationary and rotating parts

the radial growth of the steam path parts during operation


due to temperature and stress effects

the axial pressure deflection and thermal creep deformation


(in the higher temperature stages) at the diaphragm inner
web

These conditions must be reviewed, as they occur in the steady


state condition (at all loads and maximum steam conditions), and the
rate at which they change during transient operation.

94
Steam Path Component Alignment and Stage Spatial Requirements

ROTOR VERTICAL DEFLECTION


Figure 2.3.1 demonstrates the basic principle of rotor deflection.
It represents a simple constant diameter spindle having four integral
wheels, each carrying at least one blade row. A single span rotor of
length (L) is supported on two bearings (1 and 2), one close to each
end of the span. Under the action of its own weight, the rotor will
deflect (or sag) as shown in Figure 2.3.1(b). If the rotor is of uni-
form section along its length, the form of the deflection will be a true
catenary. There are, however, factors that will influence the actual
form of the deflection:

The rotor central portion will not normally be of constant


diameter. There will be step changes at many locations, par-
ticularly if the rotor has stepped segments at sealing loca-
tions. The stiffness of the rotor depends upon the rotor diam-
eter cubed, and the deflection () is a function of this stiffness

If the rotor carries shrunk on discs (as shown in Fig. 2.3.2)


these discs will normally be of different shapesand there-
fore, of different massso the mass distribution along the
length of the span will not be constant. Also, the blades will
be of varying radial length and of increasing mass towards
the exhaust end. These wheel-and-blade masses add weight,
depending upon their total volume, but add no stiffness to
the rotor section modulus. Deflection is a function of the
central spindle diameter and mass distribution, but will be
essentially the same as shown in Figure 2.3.1(b)

Figure 2.3.1 shows a total center deflection of g. If the wheel


and blade masses shown in Figure 2.3.2 are the same, the maximum
deflection s will be equal to g. The deflection is essentially a func-
tion of the diameter (d) of the central portion of the rotor.

95
Turbine Steam Path Maintenance and RepairVolume One

Solid Gashed Rotor

D
1 T 2 C

(a)

(b)

Fig. 2.3.1Deflected form of2.3.1


Figure a simple shaft and
wheel construction.
Deflected form of a simple shaft and
wheel construction.

Rotor with Shrunk on Wheels

D
1 T 2 C

(a)

L s

(b)

Fig. 2.3.2The deflected


Figureform of a shaft with
2.3.2
shrunkTheon deflected
wheels. form of a shaft with shrunk
on wheels.

The rotor material temperature will vary along its axial length
because there is a temperature reduction in the stages; there-
fore a distribution along the axial length reducing towards
the exhaust end. The rate of temperature reduction is
dependent upon the energy dissipated within the steam path

96
Steam Path Component Alignment and Stage Spatial Requirements

stages. With temperature variation, the mechanical proper-


ties of the rotor steel will vary along its axial length, deflect-
ing by greater amounts at the higher temperatures

Many rotors are of the barrel form, as shown in Figure 2.3.3(a).


In this figure, the basic diameter (D) has been retained at the shaft
ends, but over the central span, the effective rotor diameter has been
increased from D (as shown in Figs. 2.3.1 and 2.3.2) to Do.
Therefore, the rotor diameter is increased, increasing the stiffness
and increasing its resistance to deflection. The deflection will be
reduced. As shown in Figure 2.3.3(b), the total maximum deflection
is b, which tends to be less than g.

On each of the rotors shown in Figures 2.3.1, 2.3.2, and 2.3.3


there is a coupling flange (c) located at one end, outboard of bear-
ing 2 and thrust collar T. At this coupling, the turbine rotor is to be
connected to the driven machine (normally a generator). These fea-
turesbearing 2 and thrust collar T of the rotorprovide no
increase in the section modulus, and have only little impact on the
deflected shape. However, the bearing is normally given spherical
seating within its support structure so it is able to line up with the

Barrel Rotor

Do 2 C
1 T

D (a)
Lb
L

(b)

Fig. 2.3.3 (a) and (b)Deflected


Figure 2.3.3 (a) form
and of
(b)a barrel
construction
Deflectedrotor.
from of a barrel construction rotor.

97
Turbine Steam Path Maintenance and RepairVolume One

deflected angle of the rotor at this location. In Figure 2.3.3(c) the


rotor centerline is inclined to the horizontal at an angle (). To
accommodate this inclination, the bearing is turned through the
same angle () in the spherical seating of diameter (Dsp). This bear-
ing has a white metal thickness t and is designed for a radial clear-
ance to the journal of Cl.

t Cl
Dsp
D

Cl

Fig. 2.3.3(c)A bearing show-


Figure 2.3.3(c)
ing the tilt
A bearing or the
showing in the
tilt in
spherical seating. seating.
the spherical

The turbine rotors must be coupled, or connected to the machin-


ery they drive. The majority of larger output units comprise more
than one turbine section. When multi-section rotors are coupled,
they can be connected by a flexible element or use a system con-
necting individual sections together in such a manner they perform
as (and have the characteristics of) a single integral shaftthey are
solidly coupled! All modern large rotor designs use the solid cou-
pling system. In the following discussions of vertical deflection of
multi-segment rotors, a solidly coupled system is assumed, and cou-
pling flanges are required on both ends of such shafts. The exception
is rotors located at the ends of the catenary.

Consider the three-rotor string as shown in Figure 2.3.4. There are


two turbine rotorsa high-pressure section of length (Lhp) and a dou-

98
Steam Path Component Alignment and Stage Spatial Requirements

ble flow low-pressure section of length (Llp). These are coupled to


drive a generator of length (Lgen). Each rotor is individually support-
ed between a pair of bearings (1 and 2) for the high-pressure section,
(3 and 4) for the low-pressure section, and (5 and 6) for the generator.

These three rotors are shown in Figure 2.3.4 as having small axial
gaps (G1 and G2) between their coupling faces, and each rotor has
a natural deflection, shown as calculated for the hot running condi-
tions. If, as shown in the lower portions of the figures, the deflection
is drawn with the bearings adjusted to a true horizontal position (H-
H) because of the deflected form of the rotors [with inclination of the
type shown as in Figure 2.3.3(c)], the gaps between the coupling
faces would not be parallel. Consider the gap G-1 between the high
and low-pressure elements shown in Figure 2.3.5. In this case, the
gap G-1 is the mean gap as it exists at the center position of the
rotorwhich, due to their deflection, are at angles 1 and 2 to the
horizontal as shown in Figure 2.3.6. With these inclinations, the
rotors cannot be connected and the gap closed at their flanges. Even
if the coupling studs could pull the faces together, this would induce
a high level of bending stress into the outer fibers on the rotor mate-
rial and require considerable distortion to allow the gap to become
parallel and close.

G1 G2

1 2 3 4 5 6

Lhp Llp Lgen

H H

G1 G2

Fig. 2.3.4A three section arrangement showing


Figure 2.3.4the deflections in the original setting
before bearing
A threeadjustment.
section arrangement showing the deflections in the original setting
before bearing adjustment.

99
Turbine Steam Path Maintenance and RepairVolume One

Note: In fact, the gap does not close. A spacer plate between the
coupling faces is adjusted in thickness to allow fine adjustment of
the axial distance from the active thrust face to achieve the desired
clearances in the individual sections. However, before using the
spacer plate, the rotors must be adjusted so the axial gap is G-1 at
all circumferential locations.

G1
1 2

1 2
H

Fig. 2.3.5Coupling gap2.3.5


Figure at initial set-up.
Coupling gap at initial set-up.

The method used to adjust the coupling axial gaps so they are
spaced equally at top and bottom positions is as follows:

Select one section of the total string and arrange for its bear-
ings to be horizontal

Raise bearings on one adjacent section until the gap between


its coupling flange and the horizontal section are equalized.
Figure 2.3.6 is the adjusted position on the rotors, for which
the high-pressure section bearing 1 and 2 have been raised
by amounts 1 and 2 to achieve a gap (G-1) at top and bot-
tom positions. Note the centerline of the horizontal set rotor
is not horizontal, but inclined at an angle (2) due to its own
deflected form. Therefore, bearings 1 and 2 must be raised to
turn the high-pressure rotor through an angle (1 + 2)

100
Steam Path Component Alignment and Stage Spatial Requirements

2
G1
2

1+2 2

H H

Fig. 2.3.6The coupling


Figurefaces after adjustment to
2.3.6
achieve a constant
The coupling gapafter
faces G1 at top andtobottom.
adjustment achieve a
constant gap G1 at top and bottom.

Similarly, at the other gap (G-2 between the low-pressure


section and the generator), the generator rotor must be
adjusted by raising bearings 5 and 6 by amounts 5 and 6.
The relative coupling positions are then as shown in Figure
2.3.7, where the high-pressure rotor has been adjusted so the
centerline tilt angle has been modified from 1 negative to
2 positive. The coupling face of the high-pressure rotor is
now at the same angle as the low-pressure rotor. Similar
adjustments have been made between the low-pressure and
generator rotors

The rotors of modern turbine generators are solidly coupled and


can be considered as a single shaft of varying sections with loads car-
ried at different positions along the axial length. Consider the rotor

G2 5
1 G1 6
2
H
H
1 2 3 4 5 6
Lhp Llp Lgen

Figure 2.3.7
Fig. 2.3.7The three
The threerotor unitofoffigure
rotor unit figure 2.3.4
2.3.4 afterafter bearing
bearing verticalvertical height adjustment.
height adjustment.

101
Turbine Steam Path Maintenance and RepairVolume One

shown in Figure 2.3.8(a). It comprises a single-flow high-pressure


section, a double-flow intermediate-pressure section, and two dou-
ble-flow low-pressure sections. The rotors in these sections are cou-
pled to a generator rotor at the low-pressure end. In the case shown,
each individual turbine rotor is supported by two bearings, which
between them carry the entire weight of the rotating portion (includ-
ing the overhung sections from the bearings to the coupling faces). It
is also possible in some designs that bearings will be shared between
adjacent sections. Here, two rotors share a common bearing.

If the rotor had a uniform cross section and a uniform mass dis-
tribution along its axial length, and was at constant temperature, its
deflected form would be a pure catenary. However, due to the
uneven distribution of mass and section modulus and the effects of
temperature variations, this pure form catenary is modified. Figure
2.3.8(b) shows a typical deflection curve for the solidly coupled
rotor from the horizontal (H-H). As drawn, the vertical scale has
been considerably enlarged.

In Figure 2.3.8, the five rotor sections will require 10 bearings.


Because the rotor will operate in this hot, deflected form, it is nec-
essary to adjust the bearing elevation so each will assume an incli-
nation to permit flange face gap parallelism and an elevation such as
1-10 in Figure 2.3.8(b).

During site installation, the manufacturer may adjust the relative


bearing elevations to any convenient level. A common manufactur-
ers practice is to set the low-pressure sections (where the turbine has
an interface with the condenser) at the same elevation. For multiple
low-pressure sections, the low-pressure sections are set at the center
of their span. The high and intermediate pressure and generator sec-
tions must then be set to accommodate this. Such an adjustment of
the unit rotors is shown in Figure 2.3.8(c); the elevations are adjust-
ed from the horizontal (H-H) by amounts 1-10.

102
Steam Path Component Alignment and Stage Spatial Requirements

Using bearing 7 as an example, the initial determination of this


bearing is set at a distance 7 below the horizontal. In the final
adjustment, this bearing is raised by an amount 7 to allow bearings
6 and 7 to be set at a common height at the center of the low-pres-
sure sections span. The total downward adjustment of bearing 7 is
as -7+ 7 as shown.

G1 G2

1 2 3 4 5 6

Lhp Llp Lgen

H H

G1 G2

Fig. 2.3.8The vertical alignment of aFigure five section,


2.3.8 ten bearing turbine-generator, with
bearings 6 and
The vertical 7, which
alignment of a fiveare attached
section, to theturbine-generator,
ten bearing condenser setwithhorizontal.
bearings 6 and 7,
which are attached to the condenser set horizontal.

Factors contributing to vertical movement


The final, cold vertical bearings settings are influenced by consid-
erations other than the rotor-deflected shape. These must be consid-
ered and accounted for in making the cold stationary setting. Principal
issues among these second-order considerations are following:

Bearing oil rise. The oil film thickness causes the bearing to float
the rotor vertically above the white metal surface. This bearing rise is
small but should be considered in setting the nominal bearing
height. This is discussed below.

The vacuum deflection. During operation, a vacuum exists at the


turbine/condenser interface. If the condenser is mounted directly
beneath the low-pressure sections, there will be a large net down-
ward force on the low-pressure casing, caused by the difference

103
Turbine Steam Path Maintenance and RepairVolume One

between atmospheric pressure and the vacuum produced in the low-


pressure hood. Often the bearings are formed as part of the low-pres-
sure hood and carried within it. If the bearing housings are an inte-
gral part of the low-pressure structure, then these bearings will be
subject to the same vertical forces causing a deflection of the hood,
i.e., these bearings will rise and fall with the casing. In establishing
the low-pressure section bearing vertical setting, it is necessary to
make allowance for this movement and consider the effects of part
load and back pressure variations on total elevation.

Some low-pressure section designs have their bearings support-


ed from pedestals located between the low-pressure sections. In this
case, the casing will be subject to elevation changes introduced by
the vacuum force but the rotor will be unaffected. In such designs it
may be necessary to arrange for the unit to have larger radial clear-
ances at the shaft-end positions to accommodate the vertical adjust-
ment of the casing relative to the rotor.

Hot well water. During operation, the condenser hot well will fill
with condensate awaiting removal by the condensate and boiler feed
pumps. This hot well does, to a degree, act as a reservoir; therefore
the level of water contained in the well will change subject to load,
boiler demands, and other factors. It would be expected as water
level changes, the static load produced by the weight of water in the
hot well carried by the condenser will change, affecting the height
of the condenser and the load transferred to the low-pressure hoods.

There are various methods of supporting the condenser. Some


are suspended from the turbine and given some support from below.
Others are carried on spring supports and connected to the turbine
by flexible joints. Each will have a different influence on the hood
deflection and needs to be factored into the total vertical setting of
the bearings.

Bearing temperature effect. The bearing and its support structure


are subject to temperatures above atmospheric and temperature vari-

104
Steam Path Component Alignment and Stage Spatial Requirements

ations. These temperatures can be marginally different from bearing


to bearing and are dependent upon the location within the rotor
string, the amount of heat conducted along the rotor, and the effec-
tiveness of the atmospheric cooling. Because of this effect, the bear-
ing supports will expand vertically by different amounts.

Flexibility of bearing supports. Because casings expand and con-


tract due to temperature effects, they can impose varying loads on the
bearing supports and can therefore modify their elevation. This effect
is particularly noticeable with separate bearing pedestals that must
transmit axial thrusts and at the same time maintain alignment.

Centerline support of the casings. The method of supporting the


casings above the foundation, and the height at which the centerline
of the casing is located, can have an effect on the height at which the
centerline exists under varying temperature conditions. For this rea-
son the design is supported from a point at the centerline, to the
greatest extent possible. The upper half expands upwards and the
lower half expands downwards, maintaining an equilibrium condi-
tion at the centerline.

Bearing loading. The designer can deliberately raise or lower a


bearing relative to its modified catenary position obtaining a proper
bearing loading. This is a practice used when there is a possibility of
inducing oil whip into the bearings. It is known by increasing bear-
ing loading the onset of this phenomenon will be delayed.

Bearing spherical seating. The bearings must be able to accom-


modate the angle of the journals inclination so a clearance Cl is
maintained between the bearing centerlines and the journals. If for
some reason the spherical seating of the bearing does not allow
adjustment to the correct angle [, as shown in Fig. 2.3.3(c)], then
the clearance will not be correct at all axial positions of the journal.
This can have an effect on the elevation at that point.

105
Turbine Steam Path Maintenance and RepairVolume One

It is clear from these considerations that final bearing elevations


1-10 are selected with care after an evaluation of the various
factors that can influence their operating elevation. This elevation
must be achieved with considerable accuracy during site erection
and subsequent adjustment. Some of these factors have minimal
effect on the final design elevationhowever, they do have an effect.
It is necessary for the designer to consider these, if not in each spe-
cific case, at least during the development phase of the design rules
formulated to ensure the final specified elevation will help ensure
vertical alignment is maintained at all operating conditions.

Sideways shift
In addition to the requirements of vertical adjustment needed to
achieve axial gaps of G1 and G2 at top and bottom of the rotors,
these elements must be aligned side to side along their axial length
so the rotor line is true and the coupling gaps G1 and G2 (Figs.
2.3.4 and 2.3.5) are identical side to side as well as top to bottom. It
must be recognized, however, that the true operating shape of the
rotor does not run on this true side-to-side line set in the cold sta-
tionary condition. There are one or two minor influences that will
cause some adjustment but they are of a very low level and can be
ignored for all practical purposes.

The facts of a bearing oil rise were discussed previously. If the


hot running position of a bearing is considered, it will be seen that
from the cold stationary positionwith the rotor journal center
located directly above the vertical centerline of the bearing [Fig.
2.3.9(a)]there will be a positional shift as the unit goes into oper-
ation. In operation, an oil wedge is formed between the bearing
liners metal surface and rotor journal, so the rotor will in effect float
on this oil film. This film is not of uniform thickness and will be
established as a wedge [Fig. 2.3.9(b)].

106
Steam Path Component Alignment and Stage Spatial Requirements

(a)
v

(b) B

h A

Fig. 2.3.9The
Figurecenter
2.3.9line shift
dueThe
to the oil wedge.
center line shift due to the
oil wedge.

In the hot running position, the rotor has a center shift from A
to B. This shift can be considered to comprise two components
v in the vertical direction (discussed earlier), and a shift of h in
the horizontal direction. The vertical rise is considered in setting
bearing elevation. Normally, this vertical rise is greater than the hor-
izontal side shift. This horizontal shift must also be reviewed in terms
of adequacy of the units total rotational stability. The h deflection
is a relatively small amount, but where bearings of considerably dif-
ferent diameters are adjacent to each other, this can have an effect
on the oil wedge shape of both, and must be considered by design
in establishing cold clearances.

107
Turbine Steam Path Maintenance and RepairVolume One

DIFFERENTIAL EXPANSION
During operation, the stationary and rotating components of the
turbine absorb and reject heat energy from the steam and expand
and contract with the temperature changes this heat introduces. The
rotating and stationary portions accept and reject this heat, both in
different amounts and at different rates, depending upon their total
mass and the extent to which they can radiate this heat to other
media. Because these stationary and rotating components expand
and contract by different amounts and at different rates, there will be
a relative positional shift between them from the cold stationary
position during normal operation. This difference becomes even
more pronounced during transient conditions, when dramatic tem-
perature changes and (in emergency conditions) heat generation can
occur within the steam path due to frictional heating caused by rotor
rotation in a partially evacuated or stagnant steam atmosphere.

Because they are anchored to the turbine foundation block at var-


ious locations, the casingsnormally one for each sectionwill
expand in an axial direction away from these anchor points. The
extent of this expansion in any one casing is influenced only by the
amount of heat absorbed, rejected, and retained by individual cas-
ings. The extent of expansion in any one casing is unaffected by other
casings and components of the unit.

The rotating portions are solidly coupled and have a total move-
ment from their one anchor point. The anchor point is the thrust
bearing and the contact point at the active thrust face. The position
of the thrust bearing is chosen to minimize the differential expansion
in the higher condition stages, where the effects of differential
expansion have, or can be made to have, less effect on the perform-
ance of the unit.

Figure 2.4.1 is a simple, single-section, eight-stage turbine. This


unit has a thrust bearing axial location T, as indicated. At this posi-

108
Steam Path Component Alignment and Stage Spatial Requirements

tion there is contact between the rotating (thrust collar) and stationary
(thrust bearing plates) components. This position represents the only
point within the unit where there will be no relative axial movement
between the stationary and rotating components as a consequence of
temperature changes. In this unit, steam is admitted through a control
stage C1. This position within the unit represents the highest tem-
perature the metal will experience. The steam will expand within the
steam path through the remaining seven stagessteam will flow to
the right (indicated as +A) and will also leak through the sealing sys-
tem installed at the shaft end (indicated as -A).

Thrust Steam Flow Direction


Block Coupling
Active Gap
Face

C1 2 3 4 5 6 7 8
T
Absolute Linear Expansion

Rotor
Expansion

Differential
Expansion
Casing ax
Expansion

-A +A
Axial Length or position

Fig. 2.4.1The differential expansionFigure


in a single section containing eight stages.
2.4.1
The differential expansion in a single section containing eight stages.

Note that expansion to the right of the thrust bearing has been
designated as positive movement and expansion to the left as nega-
tive movement. This is a convention used only to explain the effects
of differential axial movement.

Consider the effects of the steam temperatures on the individual


stages of the steam path. The control stage (accepting the inflowing

109
Turbine Steam Path Maintenance and RepairVolume One

steam) is the highest temperature region that will expand (by an


amount in the positive direction). Figure 2.4.1 shows the slope of the
control stage expansion curve will be the steepest of the individual
stages. In the stages to the right of the control stages, temperatures
will be lower and therefore the rate of expansion will be less (but still
in a positive direction). The extent to which high temperature steam
will expand through the rotor end seals will influence the tempera-
ture in this region. To the left of the thrust bearing, heat will be con-
ducted down the rotor material causing a negative movement of the
rotor.

Figure 2.4.2 shows the control stage, where steam is admitted to


the steam path. It enters the rotating blade at a temperature Ti,
expands through the rotating blades to a temperature To, and gives
a mean blade temperature Tb, which is a function of both Ti and
To.

Ta da1

Tc

da2
Ti Tb To

Te
Tr
Tre

Figure 2.4.2
Fig. 2.4.2The temperature variations in a
stage
Thewith some level
temperature of reaction.
variations in a stage with
some level of reaction.

110
Steam Path Component Alignment and Stage Spatial Requirements

The rotor wheel temperatures are similar to the blade vane on its
inlet and discharge faces. These temperatures will influence the rotor
body, together with any heating effect of the steam expanding past
the nozzle box. The temperature will range from Ti to Te. The
total result of this heating is for the rotor to expand away from the
thrust bearing by an amount depending upon the mean temperature
and the axial span of the stage.

The casing inner surface will also heat to about the same tem-
perature as Tb. The casing, however, has an outer surface; it is
insulated but it will lose heat to the atmosphere surrounding it, and
will then be at a somewhat lower temperature (Ta). This means that
the mean temperature in the casing will be lower than the mean
effective temperature of the rotating portions (considering the cool-
ing that takes place at the outer wall). The casing temperature (Tc)
is a function of both Tb and other temperatures achieved by the
rotor material.

The nozzle plate will be exposed to temperatures higher than


Titemperatures equal to the inlet temperature of the steam. Any
steam chamber within the casing will also achieve inlet temperature,
which will tend to raise the temperatures of the stationary portions
by some small amount in this region. They will still suffer the effect
of heat loss from the outer surface. In many designswhere a dou-
ble casing design is used, for instancethe outer wall temperature
of an inner casing (which locates the stationary blade rows) is the
temperature of the steam surrounding that casing.

With these changes in the relative axial position of the stationary


and rotating components, the axial clearances (set in the cold axial
position) are modified by this differential expansion. Two hot running
axial clearances are of importance in determining the initial settings
of the individual stages in the unitda1 and da2, shown in
Figure 2.4.2. These will be discussed in greater detail later.

111
Turbine Steam Path Maintenance and RepairVolume One

Consider again the single casing unit in Figure 2.4.1. This unit
comprises a central spindle with a control and seven impulse stages.
At normal operating conditions the rotor and casing will expand
from their anchor points in the rotor in the positive direction from the
thrust bearing active face. The rotor has its highest temperatures at
the front end, and so the growth in axial length will occur at the
greatest rate at this location with the rate of growth decreasing as the
exhaust end is approached. In this unit, the casing is also anchored
at the thrust-block location and will also expand in the positive
direction; however, because it is slightly cooler (losing some heat
through its outer surface), it will not expand to the same extent.

The actual location of the thrust bearing at the front end of the
unit is also important. Often this will be contained in a pedestal sup-
porting the bearings and possibly the control mechanisms. The cas-
ing may or may not be keyed to and supported from this pedestal; it
depends upon the axial location of the anchor, and is a function of
the arrangement selected by design.

The curve of total rotor expansion will be of the form shown.


Similarly, the casing will expand in the same direction, but to a less-
er amount because of its lower mean temperatures. The difference in
expansion occurring between these two curves is termed the differ-
ential axial expansion.

There is also some expansion from the thrust bearing active face
in the negative direction. The temperature of this portion of the rotor
(to the left of the thrust block) is established by heat conducted along
the shaft from the central portion of the unit. Similarly, the casing
will expand in a negative direction, again at a lower rate than occurs
in the rotating portions of the unit.

Figure 2.4.3 shows the five-section unit previously shown as


Figure 2.3.8in this case, the thrust bearing axial location (T) has
been added. In this design, the thrust bearing is located at the bear-
ing pedestal between the high and intermediate pressure (reheat)

112
Steam Path Component Alignment and Stage Spatial Requirements

sections. Also shown in the lower portion of this figure is the axial
movement from the thrust bearing of the solidly coupled rotor and
the stationary portion from their anchor points to the foundation. At
the thrust bearing axial position, the rotor has zero axial movement
relative to the stationary elements. The casings for the high and inter-
mediate pressure sections are normally located from the
pedestal/bearing support. They often have a support platform built
onto them to locate the stationary portions on either side of them.

Thrust
block 'T'

HP Reheat LP A LP B Generator
Zero axial movement line

Coupled
Differential expansion is rotor
equal to the vertical distance expansion
from the rotor to the casing
expansion lines

expansion
Total rotor
Negative
movement Positive
movement

HP casing

Reheat Generator
Ac
Ae casing LP-A casing LP-B casing casing
Ar
Anchor points

Fig. 2.4.3Indicating the differential expansionFigure 2.4.3 in a five section turbine-generator,


Indicating the differential expansion in a five section turbine-generator, showing the rotors
showing the expanding
rotors expanding from the thrust bearing, and the casings from their
from the thrust bearing, and the casings from their anchor points.
anchor points.

Using the same nomenclature in this figure as in Figure 2.4.1, it


can be seen that movement (or expansion) to the left (away from the
thrust bearing) is termed negative (for convenience), while move-
ment to the right is termed positive. Because of their locations,
both the high-pressure rotor and casing expand away from the thrust
face to the left of the thrust bearing active face. Both are therefore
considered to have a negative movement. Figure 2.4.3 shows the
total movement of the rotor begins at a thrust face (which is the zero
point), whereas the casing initial (zero) point is located at some small
distance to the leftpossibly considered the support point, since it

113
Turbine Steam Path Maintenance and RepairVolume One

is at the outer surface of the bearing pedestal. In fact, these small dis-
tances have little influence on the differential movement, but it is
wise to be aware of their existence.

An examination of the high-pressure section expansion charac-


teristics makes it clear that modified axial clearances exist between
the hot rotating and cold stationary settings during normal operation.
If a stage in the high-pressure section is examined (Fig. 2.4.4), the
axial clearances will modify. In this figure, clearances from the sta-
tionary to rotating blade at the steam path are shown as Cao and
Cai at the outer and inner radial locations, respectively. As heat is
admitted, both rotating and stationary elements will expand to the
left. Because the stationary portions will have a lower mean temper-
ature than the rotor, they will move by a smaller amount, and the
clearances Cao and Cai will increase by the amount of the dif-
ferential expansion at the location of transition from the stationary to
rotating rows. The clearance Ci (between the stationary portions
and the preceding rotating portions) will decrease by the amount of
the differential expansion at that axial location.

Another factor to be considered is the minimum clearance


between the diaphragm and adjacent wheels along their common
face (Cxi and Cxo). This clearance normally occurs at the inner
diameter position and will be considered in greater detail later.

If labyrinth seals are located at the inner diameter of the


diaphragm and are of the hi-lo type, their axial position detail
(shown by the dimension g and h in Fig. 2.4.4) must be moni-
tored. Ultimately, there could be rubs of the seal strips on the rotor
castellations that will normally destroy the seal strips. This will
degrade efficiency and stage output. These considerations also apply
to the reheat section, except the expansion of the rotor and casing
are in the positive direction.

It is clear from a consideration of the low-pressure rotors (Fig.


2.4.3) that the accumulated axial movement of a solidly coupled

114
Steam Path Component Alignment and Stage Spatial Requirements

Cao

Cai Ci

Cxi Cxo

t
h
g

Fig. 2.4.4Showing
Figurethe axial clearances
2.4.4
which shouldthe
Showing beaxial
considered in setting
clearances which
the cold stationary
should positions.
be considered in setting the
cold stationary positions.

rotor system represents considerable movement in the positive direc-


tion. The outer casings of these sections connect to external portions
of the foundation at some suitable point. Figure 2.4.3 shows how this
occurs for both the A and B sections at their casing centerlines
(position of steam admission). This is a convenient location, as it
imposes the minimum adjustment requirement on any crossover/
around piping that transports steam from the reheat section exhaust
to the low-pressure sections inlet. However, there are designs that
anchor the low-pressure outer casing at the point nearest the thrust
bearing.

As an example of low-pressure section clearance requirements,


consider the first stages of the double flow low-pressure section
LPA shown diagrammatically in Figure 2.4.5. In the cold stationary

115
Turbine Steam Path Maintenance and RepairVolume One

condition axial clearances Cat and Cag stand between the sta-
tionary and rotating blade rows at the turbine end and generator [Fig.
2.4.5(a)].

Steam
Inlet
Steam Flow Steam Flow

Cat Cag Generator


Turbine End.
End.
Rotating Stationary Stationary Rotating Cold
row row stationary
row row condition
Section Center Line.

(a)

-dats +dagr
+datr
+dags

Normal hot
Rotating Stationary Stationary Rotating running
row row row row condition

(b)
Caot Caog
Caot = Cat - datr - dats Caog = Cag - dagr + dags

Fig. 2.4.5Showing the change Figurein axial2.4.5


position from the cold stationary posi-
Showing
tion to the hot the position.
rotating change in axial position from the cold stationary
position to the hot rotating position

When the unit reaches steady operating temperatures and has


adjusted to its normal axial operating position [Fig. 2.4.5(b)], positive
movement occurs in the rotating components in both flows. The total
axial movement at this point is the sum of the movement in the
reheat section from the thrust bearing to the LPA coupling, plus any
movement from the reheat/LPA coupling to the stage point being
considered. This can be defined as an amount +datr at the turbine
end flow, and dagr at the generator end flow. At the turbine end,
the stationary blade row will have a negative movement by an
amount, -dats towards the thrust bearing (the casing being
anchored at its centerline). These two operational movements will

116
Steam Path Component Alignment and Stage Spatial Requirements

cause a modification of the turbine end clearances. Therefore, the


normal hot running clearance at the turbine end Caot is given by:
Caot = Cat datr dats

Similarly, at the generator end, there is a positive movement


(dagr) of the rotating portions of the unit, and another positive
movement of the stationary components (dags). This produces a
hot running clearance (Caog), which is given by:
Caog = Cag + dagr dags

In the second low-pressure section (LPB), there is an even


greater axial movement of the rotor since it includes the total axial
expansion of the LPA rotor, but the reasoning in establishing these
cold axial clearances is identical to that discussed for section LPA.

In the generator, the cold setting must be adjusted to recognize


this total rotor expansion. The axial clearances in the generator are
not quite as criticalattention must be given to the hydrogen seals
and the effects of differential expansion on performance are mini-
mal. There is a magnetic center to the flux field but its axial position
is not critical, and can be accounted for in the cold setting.

The thrust bearing can be mounted at various axial locations. The


most appropriate for any design is dependent upon the steam path
or section configuration. A common location for the thrust bearing
in multi-section units is between the high and reheat (or the inter-
mediate pressure) sections. This gives both sections the advantage of
minimal differential expansion and therefore the ability to maintain
close axial clearances in both sections. This is done to minimize
axial movement in these higher-pressure stages, regions where leak-
age would have the more significant degrading effect on efficiency.
The cold stationary axial clearances at any position must be set
according to the differential axial movement predicted at the loca-
tion being considered.

117
Turbine Steam Path Maintenance and RepairVolume One

The casings of the high and reheat sections are not connected, so
both must have independent anchor positions. A similar logic
applies to low-pressure casings that are anchored to minimize axial
thrusts on crossover pipes, or in the case of nuclear units, on the
valve system at the low-pressure inlet.

V V
Thrust
bearing

T Generator Exciter
DFLP" DFLP"

(a)

"1" "2" "3" "4" "5" "6" "7"


Rotor
ax

Casing

+ve movement
-ve movement
Casing anchor point of casing,
just outboard of the number "3" bearing.

Figure 2.4.6
Fig.The
2.4.6The differential expansion of a geothermal unit comprising only low pres-
differential expansion of a geothermal unit comprising only low pressure sections. This unit has
sure sections.
a rotor withThis unitbearing
a thrust has aatrotor with
the front a thrust
pedestal, andbearing at the front
the interconnected lowpedestal, and the
pressure casings
interconnected low pressure
are attached casings
to the are at
foundation attached to the
a point near foundation
the number at a point near the
3 bearing.
number 3 bearing.

Figure 2.4.6 shows an arrangement suitable for a low-steam


condition multi-flow unite.g., a geothermal application, where
multi-flows are required for the larger outputs. In this design, the
thrust bearing is located at the front end of the unit, typically in a
pedestal containing the control mechanisms and the number 1
bearing. The LPA casing is anchored to the foundation in the
region of the number 3 bearing, but just outside the casing region.
The casings (normally fabrications) are interconnected through a

118
Steam Path Component Alignment and Stage Spatial Requirements

shared bearing pedestal, containing both bearings 2 and 3. It


can be seen that there is considerable negative movement of the
LPA casing, and the same degree of positive movement in the
LPB casing. Because of the lower temperatures in these units, the
actual amount of differential is small, and this design splits the total
movement so both double-flow sections can employ small axial
clearances. The generator and exciter are shown in this figure but,
as in the case of the unit shown in Figure 2.4.3, these are not too
sensitive to differential movement and can be set with a reasonable
degree of latitude.

Transient operating conditions


In addition to the requirements of normal operating clearances, it
is necessary to consider thermal transients that can occur in the work-
ing fluid during operation. Under transient conditions, stationary and
rotating portions of the unit are subject to steam temperature changes
and accept and reject heat energy. This change in heat energy levels
occurs at different rates between the various components and influ-
ences the relative positions they attain.

Considerations of transient conditionsand resulting expan-


sionsare of considerable importance during the design phase and
the setting of cold stationary clearances. They affect the minimum
and maximum differential clearance required to ensure that rubs
between the stationary and rotating components do not occur dur-
ing both normal operation, start up, and shutdown, when both short
or cold and long or hot rotor conditions occur. The long rotor
condition occurs when the rotor temperature is above those of the
stationary portions of the unit by an amount greater than the normal
operating temperature differential. Short rotor conditions exist
when the rotor temperature is below that of the stationary portions
by more than those achieved in normal operation.

119
Turbine Steam Path Maintenance and RepairVolume One

Short and long rotor conditions occur during transients that


are most often created by start up, shutdown, or when steam is being
admitted to or cut off from the steam path. However, these condi-
tions can also be introduced by boiler temperature excursion, when
there are load changes, or in an emergency, when load is rejected
from the generator and the main stop valves close.

These various transient phenomena introduce an imbalance or


difference in steam path temperatures from normal. Then, because
of the difference in thermal inertia between the rotating and station-
ary portions of the unit, they will accept and reject heat at different
rates, expanding axially by different amounts to the extent the run-
ning clearances must be modified from normal operating values.

In the following analysis the following symbols and suffixes are


used:

C axial clearance
d axial distance adjustment
a normal design operating position
o off design operating position
t turbine end of unit
g generator end of unit
n normal temperature condition
h hot temperature condition
c cold temperature condition
r the turbine rotating portion
s the turbine stationary portion

Long (hot) rotor. Occasionally, the steam path rotor temperature


exceeds the temperature of the stationary portions by an amount in
excess of the normal operating temperature differential. In such cases,
there will be a period of time during which the differential expansion
of the rotor relative to the stationary portions of the unit is larger than
normal. Under these circumstances steam path portions of the unit
must be examined to ensure there is no reduction of hot running

120
Steam Path Component Alignment and Stage Spatial Requirements

clearances to the extent interference will occur. Interference would


cause rubs between stationary and rotating portions of the unit.

Figure 2.4.7(a) shows a three-section turbine with high, reheat,


and double-flow, low-pressure sections (the generator has not been
included). Under normal operations, the differential expansion posi-
tions are shown, with the total differential expansion at the end of
the high- and low-pressure sections indicated as axhpn and
axlpn [Fig. 2.4.7(b)]. During periods of transient conditions
when the rotor achieves a temperature in excess of the stationary
portions by a larger differential temperaturethere will be changes
in the relative positions between the components.

In the high and reheat sections, this results in an increase in the


axial clearance between the stationary and rotating components
Cao and Cai (Fig. 2.4.4) and a decrease in the clearance Ci
between the rotating blade discharge and the inlet to the following
stationary row. Normally, the clearances represented by Ci are sig-
nificantly larger than the Ca clearances, so this is rarely of con-
cern. However, in high temperature, high-pressure stages, it is
always prudent to ensure rubs will not occur, particularly at the
inner diameter of the diaphragm (clearances Cxi on the inlet and
Cxo on the outlet sides of Figure 2.4.4), where there could also be
both elastic and plastic deformation of the web.

The differential expansion changes give total differentials of


axhph at the high-pressure end and axlpn at the generator
end of the double flow low-pressure section [Fig. 2.4.7(c)].

There are other considerations in the double-flow low-pressure


section. While axial clearances between stationary and rotating rows
will increase in one flow, in the other they will also change as a con-
sequence of the thermal differencebut the direction of the result-
ing movement will be reversed. It should also be noted that the nor-
mal differential expansion in the low-pressure sections are large

121
Turbine Steam Path Maintenance and RepairVolume One

because of the accumulated effect of rotor expansion and the intro-


duction of such transients can aggravate the overall situation.

Figure 2.4.7(d) shows a diagrammatic representation of two


stagesthe first stages in the two flows of the double flow low-pres-
sure section. We assume the unit is running at steady conditions
rather than being subjected to start-up transientsi.e., it shows the
normal running clearance of Caot at the turbine end and Caog
at the generator end flows, as defined by the equations on page 117.
The values shown and defined in Figure 2.4.5 are the normal hot
running clearance.

Under conditions of the hot rotor transients, the turbine end will
experience expansion. Because of the hot rotor condition, there will
be an additional rotor growth (dhrt). Similarly, the stationary por-
tions will expand in the negative direction (from a centerline posi-
tion) by an amount -dhst. These movements will give a final hot
transient running clearance from Caot to Coth of:
Coth = Caot dhrt dhst

This analysis shows that during the long (hot) rotor condition, the
expansion at the low-pressure section turbine end causes a decrease
in the axial clearance of (dhrt + dhst), i.e., the rows move, closing
the axial clearance.

Similarly, at the generator end of the low-pressure section, the


turbine rotor will have an additional positive movement (dhrg),
and the stationary portions will have expanded by an additional
amount (in the positive direction) of dahg. This means the normal
hot running clearance (Coag) will have changed. This expansion
gives a hot transient running clearance Cogh of:
Cogh = Coag + dhrg dhsg

This analysis shows that during the long (hot) rotor condition, the
expansion at the generator end of both stationary and rotating por-

122
Steam Path Component Alignment and Stage Spatial Requirements

HP Reheat DFLP

(a)

"n" indicates normal rotor conditions Rotor


"h" indicates hot rotor conditions normal
axlpn

(b)
Casing
normal
-ve movement 0 +ve movement

axlph
axhpn Rotor
hot

Casing
hot
0 (c)

axhph Expansion
-ve +ve
Steam Inlet
and LP anchor
Thrust Bearing

point
Hot steam flow Hot steam flow
Caot Caog
Section Center Line

(TE) (GE)

Rotating Stationary Stationary Rotating


Normal row row row row

Hot

dhrt dhsg dhrg


-dhst
Coth Cogh

Long Rotor Axial Adjustment


(d)

Fig. 2.4.7The long or hot Figure


rotor.2.4.7
In (b) is shown the normal differ-
The long
ential or hot rotor.
expansion, In the
in (c) (b) iseffect
shownofthe normal differential
temperature expansion,
reduction, and inin (c)
(d)
the effect of temperature reduction, and in (d) the modified operating
the modified operating position on the first stages in the low pressure
position on the first stages in the low pressure section.
section.

123
Turbine Steam Path Maintenance and RepairVolume One

tions is in the positive direction. However, the rotor will expand by


a larger amount, with the final effect being the clearance will
increase by an amount (dhrg - dhsg).

The total expansion effect at the two extreme positions of the


unit can be noted as follows:

high-pressure end, total negative movement changes from


normal (axhon) to hot (axhph)

low-pressure extreme, from normal (axlpn) to the hot


condition (axlph)

Note: In the analysis discussed aboveand for the cold rotor


condition belowthe unit arrangement is as shown in Figure 2.4.7,
with the thrust block and low-pressure section anchors as shown.
There can be a number of locations for both arrangements, and each
unit should be analyzed for the particular arrangement used.

A similar analysis can be made for transients that occur when the
rotor is cooled by the admission of steam, which cools it faster than
the containing stationary portion.

Short (cold) rotor. When the steam temperature reduces sud-


denlysuch as load rejection with valve closureor when the unit
is shut down, two situations may result:

Unit speed does not reach emergency values (<110% nor-


mal) and the main stop valves do not close. In this situation,
steam continues to be admitted to the unit using the control
valves to lower overspeed in a controlled manner. The unit is
then reconnected to the grid as soon as it can be synchro-
nized. Under these circumstances, there will be some reduc-
tion in internal temperatures gradients and a quenching of
the materials of the steam path; however, with the readmis-
sion of steam, it is unlikely there will be significant metal
temperature changes or that mean metal temperatures will

124
Steam Path Component Alignment and Stage Spatial Requirements

drop to the extent there will be significant change in the axial


expansion and location of the steam path parts

An emergency overspeed condition is reached, i.e., the over-


speed transient reaches emergency values (>110% normal) or
the rotor accelerates at a rate allowing emergency conditions
to be reached. Under these circumstances, the main control
valves will close and the rotor speed will be reduced. Little
steam is admitted to the unit before the valves close and unit
speed will reduce until it can be placed on turning gear.
During such an overspeed transient, a number of phenomena
have the potential to affect the temperature of the fluid
remaining in the unit. These include sections containing
deposited moisturethis moisture will flash to steam as the
pressure decays. There can also be frictional windage heating
of the steam remaining in the sections, causing temperatures
to increase to values much higher than normally experienced

Under circumstances of sudden cooling, the rotorsof smaller


thermal mass and immersed in the cooler steamwill lose heat
faster than the casings; rotor temperatures will tend to be lower than
those of the casing. When this occurs, differential expansion curves
will be modifiedfrom those shown in Figure 2.4.8(a), with a hot-
end differential expansion of axhpn and a low-pressure end dif-
ferential expansion of axlpnto those shown in Figure 2.4.8(c),
with a hot-end total differential expansion of axhpc, and a low-
pressure end differential expansion of axlpc.

Note: In low-pressure sections, design geometries can be select-


ed so rotors will be of larger thermal mass, and cool slower than the
casing, which can be a simple fabrication subjected to ambient tem-
peratures. This must be considered in establishing the possible rotor
conditionsparticularly for units with multi-flow exhausts.

The short rotor condition will modify the clearance between


the stationary and rotating blade rows shown in Figure 2.4.8(d).

125
Turbine Steam Path Maintenance and RepairVolume One

HP Reheat DFLP

(a)

"n" indicates normal rotor conditions


"c" indicates cold rotor conditions
Rotor
normal
axlpn

(b)
Casing
normal
-ve movement +ve movement.
0
axhpn C
Casing
R cold axlpc

Rotor
cold Rotor
cold
0 (c)
axhpc Expansion
Casing
cold -ve +ve
Steam Inlet
and LP anchor
Thrust Bearing

Cold steam flow point Cold steam flow

Caot Caog
Section Center Line

(TE) (GE)

Rotating Stationary Stationary Rotating


Normal row row row row

Cold
dcrt
dcst dcrg
dcsg
Cotc Cogc
Short Rotor Axial Adjustment
(d)

Fig. 2.4.8The short or cold rotor.


FigureIn (b) is shown the normal differential
2.4.8
expansion,
The short or in (c) the
cold effect
rotor. In (b)of
is temperature reduction,
shown the normal andexpansion,
differential in (d) theinmodified
(c) the effect of
temperature
operating position on the reduction, and in
first stages in(d)
thethelow
modified operating
pressure position
section.
on the first stages in the low pressure section.

126
Steam Path Component Alignment and Stage Spatial Requirements

These normally run clearances Caot at the turbine end and Caog
at the generator end and will be modified as shown to Cotc and
Cogc at the turbine and generator ends respectively. Depending
upon the rate at which the rotor cools relative to the casing, it is pos-
sible for the rotor to be at a lower temperature than the casing.

At the turbine endand under conditions of a cooled rotor


there will be a rotor contraction towards the thrust bearing of dcrt;
similarly, the stationary portions will contract, but move in a positive
direction, towards the section center (anchor) by an amount dhst.
These movements will give a final cold transient running clearance
(Cotc) as follows:
Cotc = Caot + dcrt + dcst

This analysis shows during the short (cold) rotor condition, the
contraction at the turbine end causes an increase in the axial clear-
ance at the turbine end of (dcrt + dcst), i.e., the rows move, increas-
ing the axial clearance.

Similarly, at the generator end, the turbine rotor will experience


a negative movement or contraction of dcrg towards the thrust
bearing, and the stationary portions will have contracted by an
amount dcsg in the negative direction towards the section center-
line. Therefore, the normal hot running clearance (Caog) will have
changed. This contraction gives a cold transient running clearance
(Cogc) of:
Cogc + Caog dcrg + dcsg

This analysis shows during the short (cold) rotor condition, the
contraction at the generator end of both stationary and rotating por-
tions is in the negative direction. However, the rotor will contract by
a larger amount, with the final effect being the clearance will
increase by an amount (dhrg - dhsg).

127
Turbine Steam Path Maintenance and RepairVolume One

The design process considers and evaluates these long and


short rotor conditions based on assumptions of the possible maxi-
mum and minimum temperature excursions that can be anticipated.
This excursion review establishes minimum cold stationary clear-
ances so there will be as few rubs as possible during operation.
In addition to considering the clearance effect within the steam path
stages, it is necessary to consider other locations within the unit
where clearances are tight and rubs or excessive movement could
affect total unit operation. Such locations can include oil seals adja-
cent to bearings, particularly at the low-pressure section where oil
could be drawn into the condenser; also, at barring gear locations
where misalignment of the gears can occur. Each of these axial posi-
tions needs evaluation and the cold clearance set, recognizing the
damage potential these locations represent under the influence of
transients.

The effects of thrust reversal


The steam path utilizes a high-energy fluid at various pressure
levels that exert a thrust on each surface with which they are in con-
tact. Around the surfaces of the rotating portions of the unit these
pressures exert axial thrusts that must be balanced. On the inner sur-
faces of the casing these pressure thrusts are contained by the hoop
stresses developed in the casings. Pressure drops also exist across
the stationary blade rows, and the pressures developed between the
blade attachments and the casings balance these thrusts. This fluid
pressure is sufficient that in the case of diaphragms they are limited
from moving in an axial direction by the force developed between
them and the casing location. However, as will be discussed below,
the diaphragms, as a consequence of this axial thrust, will deflect
elastically; in certain circumstances, in a plastic manner as well. In
the axial direction there are thrusts that need to be balanced.

In addition to the axial force developed by the steam pressure,


there is also an axial thrust developed as a consequence of steam

128
Steam Path Component Alignment and Stage Spatial Requirements

momentum change within the rotating blade rows. This thrust acts in
the same direction as the pressure declines from inlet to exhaust.

As a consequence of steam momentum and fluid pressure with-


in the casing, an axial thrust develops on each rotor, causing it to
attempt to move in the direction of these axial forces to the greatest
extent possible. To prevent (or limit) such unrestricted movement, the
unit contains an axial thrust bearing. Such a bearing must contain
faces intended to control and limit this axial migration while main-
taining axial rotor position within close tolerances. Thrusts are in fact
developed on the rotor in both directions, and the thrust bearing is
sized to control these total differential effects and minimize damage
that might result from excessive level or reversal of thrust.

If the unit is to operate as designed, the design engineer is


responsible to evaluate thrust levels developed, and then to select
design parameters to help ensure the net or resultant thrust is always
in the same direction, and of manageable magnitudes. In addition, a
margin is always designed into selecting this thrust bearing sufficient
to allow for some level of operating damage and for deficiencies to
be accommodated as they affect the thrust. Four phenomena, or
components, affect the total axial thrust.

Piston thrust. At any position on the rotor where there is a


change of section (increase or decrease in diameter), there will be an
axial thrust produced. These forces are known as piston thrusts, and
in magnitude are equal to the product of the pressure acting on the
vertical face and the exposed axial annulus area. Such thrusts are
present at both ends of each rotor and at other section changes. The
end thrusts oppose each other. By judicious design they can be
arranged to help balance thrust to a considerable degree.

Consider the four-wheel rotor shown in Figure 2.4.9(a). Here the


radial dimensions are shown as diameters D1 to D4. [For con-
venience it is assumed that the rotor body diameter (D4) is con-
stant at all axial positions]. In the details of the rotor end positions

129
Turbine Steam Path Maintenance and RepairVolume One

Steam Flow
4
2 3
1

(a) P1 P2 P3 P4 P5
G1 G2 G3 G4 Dr

+ve (a) -ve


D1 D2 D3 D4 D5 D6 D7
Pg1 Pg3
G1 P1 P5 G3
G4 (c)
G2
Pa Pa

-ve
(b)
D1 D2 D3 D4 D4 D5 D6 D7
+ve -ve

Fig. 2.4.9The Piston Thrusts developed on the rotor ends.

[Fig. 2.4.9(b) and (c)], where flow-restricting glands (G) are used,
there are changes in shaft diameters in addition to axial position
pressure changes. Consider the high-pressure end. Two gland rings
have a pressure differential across the first ring (G1) of P1 to
Pg1 and on the second ring (G2) from Pg1 to Pa. At the shaft
position, where the steam enters the first gland ring, there is an
exposed axial area of:

4
( D4 2
- D3
2
)
This annular area has a steam pressure (P1) acting on it.
Therefore, there is an axial thrust (Ta) developed that is equal to:

Ta = P1 .

4
( D4 2
- D3 )
2

Similarly, at the first leak off position between glands G1 and


G2 there is a vertical exposed area of:


4
( D3 2
- D2
2
)

130
Steam Path Component Alignment and Stage Spatial Requirements

With a face pressure of Pg1, this produces an axial thrust (Tb)


of:

Tb = Pg1 .

4
( D3 2
- D2 )
2

Similarly, at the second leak off positionsteam escaping past


gland ring G2there is an axial thrust Tc of magnitude:

Tc = Pa .

4
( D2 2
- D1
2
)
In this case, Pa is shown as atmospheric pressure being equal
to 14.7 psia. At the shaft end, another small thrust (Td) is produced
by atmospheric pressure. Its magnitude is:

Td = Pa .
2
. D1
4

Because the final leak off is to the atmosphere, thrusts Tc and


Td can be combined. However, for purposes of demonstrating
methodology, these are best considered separately.

The total axial thrust developed at the high-pressure end


(Thp+) can be found from:

Thp+ = Ta + Tb + Tc + Td

The implication of the + sign is to indicate the same conven-


tion of considerations or phenomena acting to the right are given a
positive notation, and those to the left a negative notation.

At the low-pressure end the same reasoning is applied. This gives


an axial thrust Te, in the negative direction at entry to the leak off
G3 of:

Te = P5 .

4
( D4 2
- D5 )
2

131
Turbine Steam Path Maintenance and RepairVolume One

At the leak off between G3 and G4, there is a thrust (Tf)


that again is negative, and has the value:

Tf = Pg3 .

4
( D5 2
- D6
2
)

Using the same logic, the thrust (Tg) at the atmospheric leak
off is:

Tg = Pa .

4
( D6
2 2
- D7 )
The axial thrust developed on the shaft end must be considered
in terms of the thrust developed on the coupling flange; the fact that
air exists on the coupling face (both sides), and the coupling flanges
(when joined) are of the same diameter. It is also necessary to con-
sider that the total thrust (Thp-) is negative. Therefore Thp- is
given by:

Thp- = Te + Tf + Tg

This axial thrust (-Thp) represents the total thrust acting to the
left in Figure 2.4.9(a) and (c). Taken together with any thrust devel-
oped on the coupling flange facenormally zerothe total axial
piston thrust (Taxp) developed on this rotor can therefore be found
from:

Taxp = (Thp+) + (Thp-)

The gland systems shown in Figure 2.4.9 represent a relatively


simple arrangement. In fact, the gland systems at both high- and low-
pressure ends of the rotor are considerably more complex than those
shown in Figure 2.4.9. Figure 2.4.10 is a sealing system in which one

132
Steam Path Component Alignment and Stage Spatial Requirements

leak off (between gland rings G2 and Gx) goes to a sub-atmos-


pheric gland-sealing condenser. In this case, the total thrust (Thp+)
would be determined from:

Thp = Pa .

4
( D4 - D3 ) + Pg1 . 4 ( D3 - D2 )
2 2 2 2

Px . ( D2 - Dx ) + Pa . ( Dx )
2 2 2
+
4 4

=
Pa ( D4 - D3 ) + Pg1(D3 - D2 ) Px ( D2
2 2 2 2 2 2
- Dx ) + Pa . ( Dx )
2
4

This arrangement is significant because the pressure Px is sub-


atmospheric, and has therefore a lesser effect on the total axial
thrust.

Pg

Px G1 P1
G2
Gx

Dx D2 +ve D3 D4

Fig. 2.4.10A shaft end sealing system


Figurewith a sub-atmospheric leak off at Px.
2.4.10
A shaft end sealing system with a sub-atmospheric leak off at Px.

Other considerations of total piston thrust are those related to


reverse flow steam paths, where either the high-pressure expansion
is broken into two separate and opposite flows, or where a combined
high and reheat steam path are contained on a combined rotor.
Consider the following arrangement: In the high/reheat section of a

133
Turbine Steam Path Maintenance and RepairVolume One

portion of a combined rotor shown in Figure 2.4.11, the high-pres-


sure section has a shaft diameter of D4a and the reheat section, a
shaft diameter of D4b. At the crossover positionwith a seal sec-
tion Gcthe rotor has a diameter of D8. Therefore, there is a
negative thrust (Th) of:

Th = Pa .

4
( D4a 2
- D8
2
)

Steam Flow

Pressure Balance P8
holes.
Steam Flow

P1 Drb

Dra 8 9
2 1 Gc

+ve -ve
D4a D8 D4b

Figure
Fig. 2.4.11Showing the piston effects 2.4.11 point in a high/reheat section.
at a reversal
Showing the piston effects at a reversal point in a high/reheat section.

A similar positive thrust (Tr) develops on the reheat section.


This is given by:

Tr = Pa .

4
( D4b
2
- D8
2
)

134
Steam Path Component Alignment and Stage Spatial Requirements

The resultant thrust (T) at this crossover position is equal to the


difference in the two values given by the previous equation and that
following.
T = Tr - Th

This piston thrust is a major contributor to the total thrust the


axial bearing must accommodate. By selection of rotor diameters
and arrangement of the steam path elements, this thrust can be
reduced to a value the bearing can carry without exceeding accept-
able values.

Wheel face thrust. If a pressure differential exists across a rotat-


ing blade row, there must also be a pressure difference (and there-
fore axial thrust developed) on any wheel face or vertical face that is
part of that row.
Many stages are designed with the blade elements carried on the
outer rim of wheels that are either formed integral with the central
rotor or shrunk onto it. With this structure, thrust develops across
these wheels, which act in the axial direction of steam flow and in
effect represents another form of the piston thrust. Consider the four-
stage rotor shown in Figure 2.4.9. Its four wheels each have a blade
root diameter of Dr. Total thrust developed on the wheels is the
sum of the individual wheel thrustswhich is again the product of
the pressure differential across the wheelsand their vertical face
area. For the rotor shown in Figure 2.4.9, the axial wheel thrust
(Tw) is given by:
2 2
Tw = [Dr - D4 ] . (P1 - P2) + (P2 - P3) + (P3 - P4) + (P4 - P5)
4

In the impulse design stage, the magnitude of the reaction at


the vane root section is relatively smalland normally just posi-
tivealthough negative reaction stages have been designed to meet
other requirements. This means the pressure differential across each
stage is relatively small. However, many impulse wheels have a large

135
Turbine Steam Path Maintenance and RepairVolume One

axial area, and can therefore develop significant levels of thrust in


the steam flow direction. To minimize the possibility of this thrust
causing excessive axial loads, pressure balance holes in these
wheels allow equalization of pressure, helping to minimize the axial
thrust developed across them. These holes also allow for the effects
of transients or mechanical damage that could introduce some mal-
distribution of stage pressures. This allows them to equalize them-
selves without allowing high levels of thrust to develop.

Do
T Dti Dto
P2 P3 Ds P3 +ve
P2

Dri Dro
Dr
d

D4
Dh
D4
+ve

Figure
Fig. 2.4.12The 2.4.12
pressure balance hole Figure
Fig. 2.4.13A 2.4.13
single reaction
at a diameter Dh. balance hole
The pressure A single
stage on a reaction stage on a
drum rotor.
drum rotor.
at a diameter Dh
Consider the wheel shown in Figure 2.4.12. A pressure balance
hole of diameter d has a mean diameter Dh. At this mean diam-
eter there is no pressure differential across the wheel, as pressure is
equalized by flow through it. However, at the diameter of the blade
root Dr, some small pressure differential exists due to the root reac-
tion. If this root section pressure drop is defined as dP, then a lin-
ear reduction from dP at Dr to zero at Dh can be assumed to
exist. Then the wheel thrust Twh in this case can be taken to be:

Tr = 0.5 . dP.

4
( Dr
2
- Dh
2
)

136
Steam Path Component Alignment and Stage Spatial Requirements

Another portion of this thrust that can be determined is normal-


ly of such small magnitude it is ignored for all practical purposes.
This is the thrust Tcv developed on the coverband, and is equal to:


Tcv = (P2 - P3) . . T . Ds
4

Where P2 and P3 are the pressures existing at the tip section,


they are not of the same magnitude as those present at the root sec-
tion; a vortex calculation must be performed to determine their val-
ues. The mean diameter of the coverband is Ds and its radial thick-
ness T.

In the case of a reaction stage carried on a drum rotor (e.g., the


rotating blade row shown in Fig. 2.4.13), a pressure drop develops
across the rotating blade and acts on the root sectionfrom diame-
ters Dri to D4, and at the tip or outer diameter, from Do to
Dti. If it is assumed that these pressures vary with radial reaction,
then the thrust Twr can be found from:

Twr =

4 {( 2 2
Dri - D4 ) . ( P2r - P3r) + ( Do 2
- Dti ) . ( P2t - P3t)
2
}
Note: This equation is developed from considerations of pres-
sure differentials from inlet to discharge rather than total thrusts in
the upstream and downstream direction. This equation also ignores
the effect of sloped sidewalls and the axial component of any thrust
developed on them. (See also the equations on page 138.) It also
takes into account:

the effect of blade damageif the blade rows sustain any


form of mechanical damage, it modifies the row discharge
area and affects the pressure distribution throughout the
steam path (and therefore, the axial thrust). In the majority of
incidents, this will be of no consequence; however, there
could be circumstances in which this can increase the load

137
Turbine Steam Path Maintenance and RepairVolume One

on the thrust bearing to unacceptable levels, and corrective


action must be taken

the effect of chemical depositsdeposits will often form as a


consequence of impurities carried over with the steam. These
impurities will often be deposited on the blade vanes and
will tend to close the throat. Under normal circumstances,
these deposits will build to a thickness of 0.005" to 0.010"
i.e., they will close the throat by 0.010" to 0.020." However,
circumstances can arise in which much thicker deposits will
form. This will influence the pressure distribution and there-
fore, the axial thrust

Blade annulus thrust. A pressure differential exists across the


rotating blade annulus. This totally axial force is the product of the
rotating blade row pressure drop and the annulus area across which
it acts. From the inlet to the row, the pressure is not of equal magni-
tude along the entire radial height, because of the radial flow com-
ponent of the steam velocity. However, at the discharge surface, the
pressure is sensibly constant at all radial heights. For blades of rela-
tively small radial height (height/mean diameter <0.20")as shown
in Figure 2.4.13the pressure at the mean diameter (P2 and P3) can
be used with sufficient accuracy that the axial thrust Tba on the
row can be determined from:
2
Tba = ( P2 - P3) .
( Ds-T/2 ) 2
- Dr
4

Equation 2.4.13(a) is suitable for use with an impulse blade row


(Fig. 2.4.12) with a relatively small pressure drop developed across
it. However, in the case of a reaction stage having a high degree of
reaction, it is necessary to evaluate the entire stage, including those
developed on the inner and outer sidewalls of the vane itself. Figure
2.4.13 shows a reaction rotation blade row that has sloped side-
walls. In this case the total Tba thrust consists of the following
terms:

138
Steam Path Component Alignment and Stage Spatial Requirements

Tba1 = P2

4
{( 2
Do - Dti
2
) + ( Dti 2
- Dri
2
) + ( Dri 2
-D4
2
)}

Tba2 = - P3

4
{( 2
Do - Dto
2
) + ( Dto 2
- Dro
2
) + ( Dro 2
- D4
2
)}
The outer and inner sidewalls carry a mean pressure 0.5(P2 + P3).

Tba3 = -
P2 + P3
2
.
4
( Dto 2
- Dti ) + P2 2+ P3 . 4 ( Dro
2 2
- Dri )
2

The total thrust (Tba) is therefore the sum of these three com-
ponents:
Tba = Tba1 + Tba2 + Tba3

Note: The equation contains those terms of the previous for the
wheel thrust Twr, but has assumed a constant pressure along the
radial height of this rotating blade row at both inlet to and discharge
from the row. This introduces no great error for small radial height
elements. The first equation has also ignored the sloped sidewalls
and any axial component of thrust developed on them. Examine all
equations in this section for blade annulus thrust.

For larger radial height blades, it is necessary to determine the


pressure distribution along the inlet height of the vane. From this can
be determined the total thrust, either by considering the mean thrust,
or determining the thrust over small radial sections, and adding their
individual values to determine the row total.

Steam momentum thrust. The change in momentum as steam


flows between the blade vanes is shown in Figure 2.4.14the force
diagram on a blade due to the change of steam momentum. The total
thrust component AD produces a component in the axial direction
equal to DE. This value must be determined for each stage in the
unit and added to the thrusts developed by the other mechanisms

139
Turbine Steam Path Maintenance and RepairVolume One

present in the unit. This axial force is directed downstream in the


direction of steam flow.

Fw = m. EA
Ft = m. DA
Fax = m. DE
Fw E
A Fax
Ft
D

Fig. 2.4.14The force diagram and its


Figure components on a turbine blade.
2.4.14
Vector EDTheisforce
the axial thrust
diagram andcomponent.
its components on a turbine blade.
Vector ED is the axial thrust component.

For any blade row, the thrust due to the change in steam momen-
tum can be determined from consideration of the vector diagram and
the change in the axial component of velocity that is achieved. For
the vector diagram shown in Figure 2.4.15, the change in axial com-
ponent of velocity Cax is given by

Cax = { C1.Sin 1 - W2.Sin2 }

so for the axial thrust Tdv for a stage flow of m #/sec, the
total thrust can be determined as:

Tdv = m
g { C1.Sin 1 - W2.Sin2 }

Pressure balance pistons


Reaction stages have an axial thrust component that is consider-
ably larger than equally-sized impulse blades. This means single-
flow sections normally require pressure balance pistons in order to
reduce thrusts to a level that normal-sized thrust bearings can
accommodate the loads developed in normal service.

140
Steam Path Component Alignment and Stage Spatial Requirements

U
1 2
1

W1

C2 C1
W2

Cax

Fig. 2.4.15The vector diagram


Figure of a single stage, show-
2.4.15
ing
Thethe variation
vector diagramof velocity in stage,
of a single the flow of steam
showing thethrough
variation
a stage.
of velocity in the flow of steam through a stage.

The balance piston consists of a wheel machined onto the rotor.


It is designed to operate with a pressure drop across it sufficient to
balance the portion of the axial thrust (which is in excess of what
might normally be developed) and capable of being carried by an
acceptably-sized thrust bearing. Figure 2.4.16 shows a typical high-
pressure, single-flow reaction section. The expansion consists of an
impulse control stage with a pressure drop across the rotating blades
from P1 to P2, and then a series of rotating blades with a pres-
sure reduction from P2 to P3. Piston thrusts also develop on each
end of the rotor. As a consequence of the reaction blade pressure
drops, a large thrust develops towards the exhaust end. To counter
this, a balance piston is incorporated into the design, with a gland
seal system above it (G4) to limit leakage. The thrust developed on
this piston (Tp) acts in the opposite direction to the blade reaction
thrust, and is equal to:

Tp =

4
{ 2 2 2 2
. P1. [D8 - D7 ] - P5. [D8 - D9 ] }
Using the pressure balance piston makes certain variables avail-
able to the designer to help select a suitable geometry. These
include:

141
Turbine Steam Path Maintenance and RepairVolume One

the point in the total thermal cycle to which the pressure P5


is connected. The steam leaking past this dummy piston is at
a high energy level and should be returned to the unit, prefer-
ably on the steam path at a pressure as high as possible

the dummy piston outer diameter D8. This diameter can be


varied within the range of what can be contained within the
casing. However, the larger this diameter, the greater the
leakage area, and the greater the energy wasted

the selection of diameters D7 and D9. These are the


diameters controlling the piston thrust area around the
dummy, and can be varied within limitations set by stress and
critical rotor speed, to control the available area

The total axial thrust developed on the rotor is the sum of these
four individual components each of which, for normal operating
conditions, is determined as part of the design process. Where pos-
sible, parameters and overall steam path geometry are selected to
minimize the required size of the thrust bearing. Unfortunately, it is
possible that during operation, transient situations can occur to
cause an increase in or reversal of the normal thrust. In the majority
of situations, this will only be a short duration phenomenon, which
will correct itselfbut during an incident, it will be necessary for the
thrust bearing to accommodate the change. While the majority of
these transient situations are of short duration, conditions can arise
that result in permanent damage to the blade system introducing a
permanent change in the thrust level. This can result in unit shut-
down so corrections can be made.

As stated, the design process determines the magnitude of the


total axial thrust of the coupled sections and arranges for a thrust
bearing of sufficient capacity to carry this load. However, in an effort
to minimize the size of the required thrust bearing, the designer will
utilize certain options to help ensure this total thrust occurs at
acceptable levels. The design options available include:

142
Steam Path Component Alignment and Stage Spatial Requirements

arranging individual sections of the unit so their axial thrusts


are in opposite directionswhere they tend to cancel

using intercept valves to shut after emergency stop valves


(ESVs). If these ESVs close too early, the high-pressure section
thrust will be unbalanced momentarily and can cause thrust
bearing failure

reversing flow in individual sections after partial expansion,


so the net thrust is reduced. This flow reversal also has other
advantages

modifying rotor diameters to increase or decrease total pis-


ton thrust effects. When making such modifications, it is also
necessary to consider the effects on the critical speeds of the
rotor

employing pressure balance pistons on reaction design


stages to counter excessive blade reaction and wheel thrust

using pressure balance holes in impulse wheels to mini-


mize wheel face thrust

arranging extraction locations in double-flow sections


where steam is extracted for regenerative feed heatingso
net thrust induced on the rotor is in the most advantageous
direction to reduce the total thrust

Thrust bearing failure


It is uncommon for a thrust bearing to failbut it can occur.
When it does:

the oil wedge between the thrust bearing and thrust collar on
the rotor will break down and cause the white metal to over-
heat and melt. This can cause extensive damage to the blad-

143
Turbine Steam Path Maintenance and RepairVolume One

ing, which, depending upon axial clearance, will make rub-


bing contact

this failure of the oil wedge allows axial movement of the


rotor, causing heavy contact between the metallic surfaces of
the thrust collar and bearing pads

A potential consequence of this contact failure is axial move-


ment of the rotorin either directionso clearances from thrust col-
lar to the active and inactive faces must be controlled; this has the
potential to modify the axial position of the rotor. It may only occur
for short periods but cause extensive damage to the rotor and bear-
ing, possibly introducing rubs that embrittle the rotor. This could
cause local hardening that may be irreparable unless it can be
machined off. This depends upon unit geometry. The blades may
require replacement.

Differential expansion with flexible couplings


The rotors discussed in previous sections have been considered
to be the solidly-coupled type, and so it is possible to consider the
total rotor string as a single, solid forging. In such a design, the
axial thrust developed in one section is transmitted through the cou-
plings to a single thrust block. It is therefore necessary to consider all
expansive movement of the rotor, in both directions, occurs from the
active face of that thrust bearing.

When flexible couplings connect rotors, the differential expan-


sion produced in one section must be carried by a thrust bearing in
that section. The thrust developed in each section must be transmit-
ted to an axial bearing located in that section, i.e., each section
requires an individual thrust bearing capable of locating the rotor
within the casing surrounding it. Therefore, conditions established by
a solidly-coupled rotor do not exist with the flexible design, and
tighter hot-running clearances can normally be accepted.

144
Steam Path Component Alignment and Stage Spatial Requirements

Modern turbine units tend toward the use of solidly-coupled sec-


tions. While this may introduce certain problems of accommodating
large axial movements, on balance these are found to provide
advantages that certainly outweigh the disadvantages.

Different forms of flexible couplings are available, and there are


advantages to each of them. If a tooth coupling is used, for example,
there is often a tendency for it to lock if transmitting higher torques
than it was originally designed for.

RADIAL EXPANSION OF
THE STEAM PATH PARTS
Steam path components exhibit growth in the radial direction
during operation. This growth is due to two separate phenomena
affecting all rotating components to a degree. One of these will also
affect the stationary components to a lesser effect. These phenome-
na are detailed below.

Thermal effects
As the unit heats with the admission of steam, the individual
components accept this heat and expand in all directions. There are
natural laws expressing the extent of expansion in terms of material
properties (linear coefficient of expansion) and the degree of tem-
perature increase. Consequences of this thermal growth in the axial
direction have been considered in the previous section, and the
modifying effects of thermal transients and other phenomena will be
considered in subsequent sections.

145
Turbine Steam Path Maintenance and RepairVolume One

In the radial direction, the effect of thermal radial expansion is


evident in both stationary and rotating portions of the unit. However,
because the casings (stationary components) are slightly cooler than
the corresponding rotating portion (due to heat radiation), the casing
will expand to a lesser extent than the rotating portions at the same
axial location.

The casings carry the stationary blades or diaphragms and radial


seal portions of the unit. These thermal effects cause the components
to expand and contract radially in response to changing temperature
environments. It should be noted that as the unit heats the casing
will radially expand outwards. Any blades or diaphragms they carry
and locate will also expand in the radial direction, maintaining their
total alignment. Stationary blades located in casing grooves will
move with the casing because they are locked in the radial direction
by the root portion of the blade. Diaphragms, however, are located
within grooves machined into the casing and will move radially, but
because they are not constrained by it, they may move at different
rates relative to the casing and their temperature ramp rates. This is
discussed in more detail below.

Radial stress effects


Mechanical stresses are induced in the rotating portions of the
unit by their own centrifugal force. These stresses have a component
that induces radial growth (or extension) in both the blades and
rotor. These stresses are proportional to the square of the speed.
Because the stationary portions of the unit are not subject to cen-
trifugal loading, these will not grow in the radial direction due to any
rotational effects. There is, however, a degree of radial growth due to
the internal pressure, causing a hoop stress in the casing walls. This
will cause some small amount of radial stress growth. Also, it is not
uncommon for this internal pressure to cause the casing to become
slightly elliptical due to the greater stiffness at the horizontal joint.

146
Steam Path Component Alignment and Stage Spatial Requirements

Stress Thermal
Thermal
expansion expansion
"drcs" expansion
"drct" "drrt"
Stress
expansion
"drrs"

"Dfs"
Cold Cold
"Dfr":
clearance blade tip Hot
rs Original diameter blade tip
Hot running cold casing diameter
clearance diameter
rh "Doc"
Cold root
diameter
Hot
blade root
diameter

Rotor Center Line


Fig. 2.5.1The thermal and stress
Figureeffects
2.5.1 influencing radial growth in a
turbine stage. The thermal and stree effects influencing radial
growth in a turbine stage.

Combined factors of radial growth


The radial growth elements for a single stage are shown in Figure
2.5.1. To the right of this figure are those effects that influence the
wheel diametral changes; to the left, those that influence the casing.

The rotating blades and rotor will grow as a consequence of both


stress and temperature. Consider first the effects of temperature on the
rotating portion of the stage. For this wheel (including the rotating
blades) the effects of radial growth are shown as drrt, due to tem-
perature. During operation, all components of this stage will reach a

147
Turbine Steam Path Maintenance and RepairVolume One

temperature that is at some variation from inlet to discharge edges,


depending upon the degree of reaction. In fact, a mean temperature
can be determined for all elements of the rotating stage (Fig. 2.4.2).
This can be taken as the mean temperature between inlet and dis-
charge, and all elements can be considered to grow radially as a con-
sequence of it. While the rotor can be considered to be at the same
mean temperature as the rotating blade, it is possible that it will also
be influenced to some small degree by heat soaking along the rotor
from the hotter to the cooler positions, and from its cooler side to the
lower temperature stages. For all practical purposes this can be
ignored.

The centrifugal extension of the rotating portions of the stage is


shown to give a total radial growth of drrs. This growth represents
the sum of the growth in the wheel and the growth in the blade,
including vane and root attachment.

Therefore, the final operating diameter of the rotating blades


(fr) is given by an extension beyond the cold stationary position
(or) by the expression:

Dfr = Dor + 2(drrt + drrs)

On the left of the stage are shown the radial changes that occur
in the casing. The original inner casing diameter is shown as Doc.
At the normal operating condition, the casing expands radially
(drct represents the total growth due to thermal influences; drcs
represents the effects of the casings radial expansion due to the
internal steam pressure). Therefore, the hot operating diameter Dfc
of the casing is given by:

Dfc = Doc + 2(drct + drcs)

148
Steam Path Component Alignment and Stage Spatial Requirements

The cold internal clearance rc is found from:

Blade tip diameter cold - Cold internal diameter of the casing


rc =
2

Dor - Doc
=
2

Taking into account the radial growth of the stationary and rotat-
ing components, the hot running clearance rh can be determined:

Dfc - Dfr
rh =
2
[Doc + 2(drct + drcs)] - [ Dor + 2(drrt + drrs)]
=
2

These clearances are calculated on the assumption that the seals


and/or wall above the blades are formed by the casing inner surface.
If this is not so and seals are carried in an extension of the
diaphragm, these equations may not be valid. This is considered in
the following section.

In the case of a casing-carried seal, the designer selects the inter-


nal casing diameter to accommodate the blade tip radial growth
with cold clearance sufficient to ensure a lack of running interfer-
ence between the parts during operation. As discussed above, the
stages can be subjected to thermal transients. They will heat various
stationary portions of the unit, expand at different rates, and be influ-
enced by the mass of the horizontal flanges that can heat and cool,
thus attaining constant temperature at a slower rate due to the large
volume of material at these locations.

In the case of low-pressure sections, the casings produced by


fabrication can heat and cool faster than the rotating portions. This
also has to be considered.

149
Turbine Steam Path Maintenance and RepairVolume One

Influence of stationary blade locating method


Stationary blade rows are located radially to provide guided flow
from the stationary to rotating blade rows. This is achieved by ensur-
ing adequate inner and outer lap in the hot rotating position. To
ensure this condition occurs, consideration must be given to the
manner in which these stationary blade vanes are located and how
temperature will affect them. There are two distinct methods of locat-
ing these blades:

The blades are located in grooves machined into the inner


surface of the casing, or a blade carrier that locates and car-
ries several rows. Upon expansion or contraction in response
to temperature changes, the blades will accept heat and
expand, so there will be no increase in gaps between indi-
vidual blade segments

The blades are located in a diaphragm produced in segments


(normally two), each of which has a 180 arc. Diaphragms
are not exclusively located in two arcs, and designs of 6-10
segments are in use

Associated with these different constructions are considerations


relating to retaining radial positions as the unit heats, expands, and
is subjected to steam loads:

For stationary blades located in a casing, the blades will


move with the casing (and therefore change their diameters)
as the casing moves radially outwards upon heating or radi-
ally inwards upon cooling

When the blades are carried in a 180 segment, they are


often bolted together at their horizontal joint, and will be less
affected by casing movement. The casing simply supplies
axial location and mechanisms to hold the diaphragm at the
correct radial location at its center. The diaphragm will
assume the lowest possible position due to its weight. If it is

150
Steam Path Component Alignment and Stage Spatial Requirements

supported at the casing centerline, weight has less influence


on radial location

Diaphragms constructed of a number of individual segments


carry the possibility of vane radial movement with tempera-
ture change. It depends upon the method used to locate the
diaphragm outer ring in the casing. Each design is considered
separately

Effects of operating transients


It has been shown that thermal transients affect axial clearances
between the stationary and rotating portions of the unit, and these
must be taken into account when setting cold axial clearances. In a
similar manner, it is necessary to consider the effect of these tran-
sients on radial settingsconsiderations influenced particularly by
the manner of construction. As in the case of axial expansion and
contraction, considerations must be made for the weight of the vari-
ous components and the rate at which they receive and reject heat
from the surrounding steam. Those considerations include the fol-
lowing:

Load rejection. The effect is triggered when the unit is discon-


nected from the grid. Under these circumstances, the rotor will be
accelerated to higher rotational speeds, increasing the radial stress
and therefore the radial extension. The designer must consider these
factors when establishing the cold setting radial clearance.

In the event of load rejection and main valve closure, both the
pressures and temperatures within the casings are reduced. In this
situation, the lower temperatures will reduce the existing thermal
radial growth of both the rotating and stationary components, and
reduce the total radial expansion effects. These diametral corrections
will not occur suddenly, but require time for the temperatures to
reach an equilibrium condition. A corresponding reduction of inter-

151
Turbine Steam Path Maintenance and RepairVolume One

nal pressure will result in a reduction in the casing hoop stress exten-
sion that will occur immediately. An internal pressure reduction will
have no effect on rotor stress extension. Therefore, as the casing
cools and contracts the rotor will increase in radial growth due to the
increase in speed. The designer setting the cold stationary clearances
must consider these effects.

Overspeed transients. When load rejection occurs, and the unit


is disconnected from the grid (or other driven device), the unit enters
an overspeed transient. The extent to which speed increases above
normal operating values depends upon a number of factorsinclud-
ing the load at the time of valve closure; the speed with which the
control and stop valves close, and (in the case of a unit with free
moisture in the various sections) the extent to which this moisture
can flash to steam with pressure decay and continue to expand
through the steam path, contributing to an overspeed increase.

There are two considerations of overspeed transients and the


extent to which these will affect the radial clearance:

The unit attains a normal overspeedi.e., the unit does not


increase its rotational speed to the extent the main stop
valves close. In this situation, steam will continue to be
admitted to the unit in such a manner that internal tempera-
tures do not rise excessively due to windage (in the case of a
main power generating unit, it will continue to carry station
auxiliaries and can be reconnected to the system as speed
reduces to normal operating values)

The unit reaches emergency overspeed values, or accelerates


at a rate that would cause a trip anticipatory device to close
the main stop valves. In this situation, clearances must be
evaluated in terms of the anticipated temperature differences
that could arise

152
Steam Path Component Alignment and Stage Spatial Requirements

Note: The trip anticipatory device is designed to sense the rate


of rotor acceleration in the event of an overspeed. If this device
determines the overspeed will reach excessive values (in excess of
normal overspeed, or about 108-110% normal), this mechanism will
trip the unit to close the main and intermediate stop valves.

In the event of overspeed, the direct centrifugal loading of the


rotating portions of the unit will increase in proportion to the speed
of rotation squared. This increase in radial stress will cause the rotat-
ing blades and wheels or rotor to expand radially. The casing unaf-
fected by this increase in speed will not increase its inner diameters.
In fact, as pressure decays (with valve closure limiting steam admis-
sion to the unit), the internal pressure on the casings will decrease,
reducing the radial expansion caused by the hoop stress induced in
the casing. In general, these diametral changes of both the stationary
and rotating components are small, but should be considered in the
design phase, particularly for the larger blade elements where stress
levels are high.

Temperature up-ramps. This pertains to some occasions when


boiler temperatures rise. During operation it is normal for the steam
to be attemperated and the temperature of the fluid entering the tur-
bine to remain sensibly constant at the design value. However, if the
temperature does manage to exceed the design values at some loca-
tion within the steam path, then local expansion will occur in the
radial direction.

Temperature down-ramps. Temperature reduction can occur in


the steam delivered from the boiler superheater or reheater. This is
not too common, and may only continue for short periods. However,
during the cooling period, both sets of components can move and
running clearance changes should be considered. This is unlikely to
be a major cause for concern relative to other factors, but in a situa-
tion where other unpredictable factors are present, it could compro-
mise running clearances.

153
Turbine Steam Path Maintenance and RepairVolume One

Water ingestion. In the event water is ingested into the unit, it


will have a quenching effect on various components, depending
upon which portions of the unit suffer the cooling. This can effect
rotor concentricity and casing alignment, both of which have the
potential to effect running clearances.

The consequences of radial growth


The design process anticipates radial growth of the stationary
and rotating portions of the unit and makes allowances for them. In
establishing the normal radial clearance, several considerations need
to be factored into the selection of the cold setting clearance, as they
will impact upon the successful running of the unit and the mainte-
nance of clearances.

Lifting of a rivet attached coverband. For those stages having a


coverband attached to the blades outer diameter by the riveting of a
tenon, centrifugal loadsinduced by their own weightact on the
coverband. These tend to make it deflect radially outward [Fig.
2.5.2(a)].

With this form of coverband attachment, the tenon will nomi-


nally hold the coverband in place, and it will be rigidly attached to
the blades outer diameter. However, under all rotation conditions,
there will be sufficient force induced by the centrifugal loading of
the coverband itself to make the inlet and discharge edges (and the
overhung coverband at the batch end edges) curl. This is shown in
Figure 2.5.2(a)a radial outwards movement of ri at the inlet and
ro at the discharge. Curling extension at batch end positions could
be even more severe.

It is not reasonable to expect this radial outwards movement


would introduce rubbing under normal operating conditions. There
is the possibility [shown in Figure 2.5.2(b)] of the coverband rising
on the tenon (particularly as the unit ages) and for total outward

154
Steam Path Component Alignment and Stage Spatial Requirements

movement to increase. This adds an additional extension (da) as


the riveting becomes less effective and increases total outward
movements to ri+da and ro+da at the discharge. Also, if the unit
goes into an overspeed conditionand particularly if the overspeed
is accompanied by higher than normal levels of rotor vibrationit
becomes possible for rubbing to occur. This may only result in wear
of the seal strips and some gouging of the coverband and does have
the potential to cause some degradation in the performance of the
unit.

da+ro

ri ro
da

(a) (b)

Fig. 2.5.2The cover band attached by riveting.


Figure 2.5.2 In (a) is shown the curling
effect
Theat the inlet
cover andattached
band discharge edges. InIn(b)
by riveting. (a)isisshown
shownthe
theeffect
curlingofeffect
time
creep in the radial
at the inletdirection.
and discharge edges. In (b) is shown the effect
of time creep in the radial direction.

Lifting of a formed integral coverband segment. If the stage con-


tains a coverband that has been formed integral with the blade vane,
then these individual segments will also curl outwards under the
influence of their own centrifugal force. However, in this case, the
amount by which these segments will deflect radially are influenced
by the following:

Form of the integral coverband. The coverband portions will


curl outwards under the effect of centrifugal loading depend-
ing upon their unsupported sections. As shown in Figure
2.5.3(a), there is an overhung portion (a) of the lozenge
shaped coverband that will tend to curl because of the dis-
tance from the profile. In this case, the extreme position (u)
is farthest from the profile and, therefore, the radially most

155
Turbine Steam Path Maintenance and RepairVolume One

deflected section moving outwards by an amount rs. It is


also that portion of the coverband most likely to rub, if there
is some surface or seal above it

Position of the center of gravity of the coverband segment. In


some designs, the shape of the coverband is dependent upon
the method of manufacture. When plunge milling forms the
vane, the coverband will also have end shapes that are con-
sistent with the profile face. Figure 2.5.4(a) shows a cover-
band segment with a joint face consistent with the pressure
face of the vane. Large overhung sections at both inlet and
discharge edges (an amount a) will induce a large curling
stress and cause these faces to move radially outwards under
the influence of rotation

Machining thickness. Many stages with integral covers are


trim machined after assembly of the blades to the rotor. This
trim machiningon surface z-z of Figure 2.5.4(b)
achieves a concentric surface suitable for sealing. It is normal
for this trimmed diameter to be determined by design, allow-
ing for a suitable leakage clearance at the tip. Should there
be an accumulation of manufacturing errors, the trimmed
thickness of the coverband (Tk) can be produced too thin.
While this will reduce the centrifugal load (and therefore the
stress) of the coverband segment, it will also cause a reduc-
tion in the section modulus of the coverband, making it more
susceptible to bending deflection. This will produce a larger
radial movement than design had intended, and could cause
excessive outward deflection. This effect of trimming is
shown in Figure 2.5.4(b).

Slanted outer sidewalls. Some stages have rotating blades with


slanted outer tip sections (Fig. 2.5.5). These stages are often used
without coverbands. In these stages, clearance must be considered
in both the radial and axial direction when the rotor is in the long
or short positions. These extremes represent critical positions, as

156
Steam Path Component Alignment and Stage Spatial Requirements

s
u

a rs
s

View in
direction 's-s'

v
(a) (b)

Fig. 2.5.3The lozenge shaped


Figure integral cover showing
2.5.3
theThe
radial outward
lozenge movement,
shaped integral under the effectthe
cover showing of radial
centrifugaloutward
loading.movement, under the effect
of centrifugal loading.

a s a s

a
Tk
z z

Fig. 2.5.4The
Figureshaped
2.5.4 cover,
and shaped
The the effect of trimming
cover, and the effect
of trimming.

157
Turbine Steam Path Maintenance and RepairVolume One

there is positive (+ve) and negative (-ve) movement. Axial differ-


ential expansion could cause the radial clearance to be consumed
and the blade tips to rub on the casing. Many designs utilize this
slanted outer surface on low-pressure stages and are most likely to
experience large axial expansions in a two-flow design. Here, the
blade tip is thinned to a squealer section to allow close clearances
between the blade tip and casing. If rubbing should occur in this
design, excessive amounts of heat are not generated, and the blade
will not bend because of the rubbing effects.

Axial
movement Long rotor
Short rotor "da" running
running clearance
clearance "Crh"
"Crc"

-v +v

Normal
running
clearance

(c) (a) (b)


Short rotor Normal Long rotor
position running position

Fig. 2.5.5The started outer side wall, 2.5.5


Figure and its effect on radial clearance.
The slanted outer side wall, and its effect on radial clearance.

158
Steam Path Component Alignment and Stage Spatial Requirements

DIAPHRAGM DEFLECTION AT
PRESSURE AND TEMPERATURE
During operation, the stationary blade rows are subjected to a
pressure differential from inlet to discharge sides. This pressure dif-
ferential is applied to impulse and reaction designed stages.
However, because the pressure drop across an impulse stage is much
higher, and because the inner surface of the stationary blade or vane
is normally attached to the inner web of a diaphragm, the conse-
quences of this condition can be much more severe on these
impulse stages. Both impulse and reaction stages can also be oper-
ating at high temperatures, which modifies the mechanical strength
of the material.

Two stationary blade cases must be considered in terms of the


amount of axial deflection: the diaphragm type stages (as used in
impulse designs) and some low-pressure stages, which have high
levels of reaction on them. There is also the need to consider the sta-
tionary blading of high reaction stages, where blade roots are assem-
bled into a blade carrier or casing.

High pressure, high-temperature diaphragms


The high pressure, high-temperature diaphragm consists of an
outer ring, located either in a groove produced on the inner surface
of the casing or on a ligament carried from the casing. Because of
this location, the outer ring cannot deflect. However, the blade row
and the inner web develop high pressures across them, and will
deflect axially.

This effect is shown in Figure 2.6.1. The pressure drop across the
vanes will deform by some amount shown as dxv and the inner
webwhich acts as a pressure barrier between stageswill experi-
ence a pressure gradient in the downstream direction deflecting the

159
Turbine Steam Path Maintenance and RepairVolume One

diaphragm towards the following row of rotating blades (which at the


inner diameter can be taken to be dxw). The total inner deflection
is therefore:
dxd = dxv + dxw

Reference point
Casing

Outer
ring
Vane
"dxv"

Inner
web

"dxw"
Total "Axial" or "dxd"
dishing movement.
Fig. 2.6.1Showing the axial deflec-
Figure 2.6.1
tion. Showing the axial deflection.

For the lower temperature, lower pressure stages, the pressure


differential is not as large. Normally the surface area exposed to the
pressure is greater, and there will be some small level of deflection.

In any stage, there will be a deflection downstream as pressure


is applied to the unit diaphragm. This initial deflection takes an elas-
tic form and will recover and return to its original axial location as
the load is removed. Long-term exposure to this pressure at an ele-
vated temperature will be sufficient to make the diaphragm creep in
the axial direction. This creep deflection is termed plastic and rep-
resents a permanent deflection. The diaphragm will not recover this
portion of the total deflection when the pressure is removed from the
unit.

160
Steam Path Component Alignment and Stage Spatial Requirements

The design-stipulated cold clearances must consider the creep


deflections and make adequate allowance in the initial setting. This
ensures that, as the unit ages and there is an accumulation of deflec-
tion, there will not be running interference. i.e., the axial clearance
Cxo (Fig. 2.4.4) will not be totally consumed and rubbing will not
occur.

Stationary blades
In the case of stationary blades, located in a casing or blade car-
rier with little or no thick inner web, there will be an axial deflection
dxv as shown in Figure 2.6.2. This refers generally to a reaction
stage with a lower pressure drop than a similar capacity impulse
diaphragm vane. However, the reaction stagebecause of space
limitationstends to utilize blades of smaller axial width, and these
tend to have less resistance to axial deformation. One manufacturer
brazes these stationary blades into groups, which supplies extra
resistance to deflection.

Casing or
blade carrier

Stationary
blade

dxv

Fig. 2.6.2TheFigure
stationary blade.
2.6.2
The stationary blade.

161
Turbine Steam Path Maintenance and RepairVolume One

A major consideration of the stationary blade is that it does not


have an inner web, and is not subject to the considerable axial thrust
that can develop across it. However, if these individual blades are
not located adequately in the casing root sloteither as a conse-
quence of poor assembly practices, poor control of tolerances on the
vane access slot width, or root tolerances not maintainedthere can
be elastic and plastic deformation, which can increase the possibili-
ty of rubs under certain conditions of blade tilt. Critical axial widths
that should be controlled are shown in Figure 2.6.3. In addition, the
caulking that is often applied to the blades must be sufficiently tight
to cause good contact between the blade root underside and the cas-
ing to prevent the blade tilting.

Ws

Wb

Cv
Cf

Wn
Cu Cu
Wc

Fig. 2.6.3The controlling


Figure 2.6.3 dimen-
sions around a stationary
The controlling dimensionsblade roota
around
and casingblade
stationary slot. root and casing slot.

Creep deformation of the diaphragms


Diaphragms that operate in the high and reheat sections of a unit
are subject to high intensity axial pressure differentials while simul-
taneously operating at high temperatures. Under these conditions, a
continual axial stress tends to deflect the vanes and inner web down-
stream. (Creep deformation is covered in greater detail in chapter 5).

162
Steam Path Component Alignment and Stage Spatial Requirements

During operation, two forms of axial deflection occurthe


deflection that is recovered when the axial pressure differential is
removed, and the deflection that is not recovered (plastic and per-
manent). This second, plastic extension is termed creep deformation.
The higher the stage temperature, the greater the degree of deflection
per unit load, and the more susceptible to this form of deformation
the element becomes.

Metals creep at all temperatures when loads of sufficient magni-


tudes are applied. However, the amount is often so low in the lower
temperature stages that this effect can be ignored. However, if tem-
perature and stresses are sufficiently high, metal will creep until rup-
ture occurs. When such combinations of stress and temperature are
severe, the creep rupture effect must be accounted for in the com-
ponent design. Creep rupture is the term used to identify the
mechanics of deformation and failure of metals under steady stress
at elevated temperatures.

The creep deflection of metals can be considered on the curve in


Figure 2.6.4. This curve is drawn to show the material deformation
(Et) that is the amount dxv and dxw of Figure 2.6.1 under a
constant load at a given constant temperature. This curve represents
a plot of material extension (or deformation) as a function of time.
The diaphragm deflection under load L falls into four distinct
stages:

Immediate extension A. Under the action of load L, this


extension is elastic and the total deflection will be recovered
if the load is removed at any time

Primary creep B. The creep continually decreases as a


result of the strain hardening the material that accompanies
the elongation

Secondary creep C. Creep ordinarily remains relatively


constant because it represents a balance between the strain-

163
Turbine Steam Path Maintenance and RepairVolume One

hardening ability of the material and its susceptibility to ther-


mal softening

Tertiary creep D. Creep is elongation to rupture, and is the


final stage. The elongation rate increases because of local
necking down, crack formation, and structural changes to
rupture at condition F

D
Et
C
B

Rupture
A "F"

Time

Fig. 2.6.4The time phasesFigure


for creep deformation to failure f.
2.6.4
The time phases for creep deformation to failure F.

The minimum creep strength of a diaphragm is defined as the


stress required to produce a specified amount of secondary creep
(C) in a given timee.g., 1% creep deformation in a time period
of 10,000 or 100,000 hours, at a particular temperature under a con-
stant stress. A materials rupture strength is usually expressed as the
stress required to produce failure, by rupture, in a stipulated time
(e.g., 100,000 hours at a particular temperature).

This creep rate is very sensitive to temperature at the higher val-


ues at which the diaphragm is operating. Families of material creep
properties are available (Fig. 2.6.5), where T1>T2 >>T5. These
curves allow the total deflection (dxd) to be predicted for a stage as
a function of its temperature, stress levels, and hours of operation.

164
Steam Path Component Alignment and Stage Spatial Requirements

The creep strength of any particular diaphragm is dependent


upon its resistance to the creep rupture effect, and the ability of the
designer to calculate its probable strength. These include:

the relative strength of the various parts of the diaphragm,


(vane and inner web)

the variation of the strength throughout the assembly due to


the changes of metal cross sections

difference in the material components, and their different


creep resistances at different positions

uncertainty of the distribution of pressure and temperature


over the component being considered

D
Et

B T5

T1

Time
Fig. 2.6.5A family of creepFigure
curves 2.6.5
for temperatures T1 to T5.
A family of creep curves for temperatures T1 to T5.

165
Turbine Steam Path Maintenance and RepairVolume One

UNIT END-TO-END
LATERAL ALIGNMENT
The total alignment of the steam path components requires they
be in-line at all operating conditions and in all directions. This
includes end-to-end alignment. To achieve alignment it is neces-
sary for the person responsible for this operation to achieve side-to-
side measurements sufficient to ensure all stationary components are
equidistant from the center on those surfaces needing interaction
with the rotating portions of the unit.

The first action step to achieving total alignment is normally to


place the bearings at the correct radial height (discussed above and
drawn for a multi-section unit in Fig. 2.3.7). The bearings are the
major consideration in the total alignment. Once they are set, it then
becomes necessary to align the other components (diaphragms,
packing heads, oil seals, etc.) so they follow the same catenary at all
axial positions along the length of the sections. Figure 2.3.9 shows
some small amount of lateral adjustment required to account for the
oil shift (dh). Normally, this is a relatively small amountbelow the
level that special provision should be made for itbut it will cause
some small adjustment in the side-to-side hot running clearances.

With bearing heights adjusted, it is necessary to adjust the


remaining components. This requires that their height first be set so
casing horizontal joints are at the correct height. Various compo-
nents are then adjusted side-to-side in their correct radial position
laterally and vertically. These components are then locked in place.

The methods of establishing the center position for gauging each


of these three parameters employ a means of establishing a straight
line from the front standard to the back-end generator or exciter. This
alignment requirement is considered in greater detail below.

166
Steam Path Component Alignment and Stage Spatial Requirements

METHODS OF
FIELD ALIGNMENT
Methods are available for achieving an alignment of the station-
ary components within the individual casings of the unit and the sec-
tions relative to each other. It is necessary to achieve this alignment
so the efficiency of energy conversion and unit reliability is maxi-
mized within unit capability.

Stationary component alignment


Within the casings there is a need to assemble the individual sta-
tionary elements (or their carriers) in the radial and axial position. (In
this section only radial adjustment is being considered.) Figures 2.8.1
and 2.8.2 show the basic settings of the steam path components in
the cold stationary position for impulse and reaction stages. (These
requirements are considered in greater detail in later sections of this
chapter.) These figures indicate the diameters and clearances that
should be achieved for the complete circumference of the blade rows
to assure a level of performance consistent with design predictions.

Figures 2.8.1 and 2.8.2 show the basic steam path diameters
(Dot and Dor) at discharge from the stationary blade rows. These
figures also indicate there is some means in the impulse stage to
make adjustment by modifying the dimension Crx to modify the
actual position of the diameters. However, this adjustment does not
modify the diameters themselves, but simply their location within
the steam path. In the case of the reaction stage in which the sta-
tionary blades are located within a casing or blade carrier, the diam-
eters are established by the diameter of the locating root shoulder
(Ds) and are not adjustable. If these blades are carried in a blade
carrier rather than a casing, they carry some level of adjustment by

167
Turbine Steam Path Maintenance and RepairVolume One

relocation of the carrier within the casing. However, any adjustment


made to one row will affect all stages located within that carrier.

Crx

Cro
Dot

Dor

Cri

Dr

Fig. 2.8.1The radial alignment


Figure 2.8.1requirements of a
wheel
The radial alignment requirementsstage.
and diaphragm constructed of a wheel and
diaphragm constructed stage.

When making any adjustment to the discharge diameters Dor


and Dot in both impulse and reaction rows, there will be an effect
on the clearance from the inner stationary blade diameter relative to
the rotor diameter Dr, which influences the radial clearance Cri.
Normally, seals will be fitted at this location with no likelihood of
rubs between the massive material pieces. However, adjustment can
cause damage to the seal strips and represent a loss in stage effi-
ciency, depending upon the sequence of assembly.

The dimensions shown in Figures 2.8.1 and 2.8.2 are for all cir-
cumferential locations and, as shown in Figure 2.8.3, must be met at
all positions.

168
Steam Path Component Alignment and Stage Spatial Requirements

Caulking
Crx

Cro

Ds

Cri
Dot

Dor

Dr

Fig. 2.8.2The radial


Figurealignment
2.8.2 requirements
of a radial
The stationary bladerequirements
alignment inserted into of
a casing or
a stationary
blade carrier.
blade inserted into a casing or blade carrier.

Cro
Dot
Dor
Dr Cri

Figure
Fig. 2.8.3The radial2.8.3
clearances and
alignment.
The radial clearances and alignment.

169
Turbine Steam Path Maintenance and RepairVolume One

Means of component adjustment


A certain level of radial adjustment is available to manufacturers
and operators of steam turbines. It represents adjustments that must
be made at initial assembly in the field, and after each component is
removed during a maintenance outage for rework, repair, or simply
for cleaning. Consider the two design concepts:

Diaphragm construction. In this concept, the stationary blades


are located within inner and outer rings and dependent upon their
method of manufacture. The values of the discharge diameters
(Dot and Dor) can be controlled within design specified limits.
Figure 2.8.4 shows two diaphragm halves located at some arbitrary
vertical distance apart (G) before closure on assembly. In the final
assembled condition they are in intimate contact. Keys are used at
their interface to locate them correctly relative to each other. How-
ever, these keys are intended and dimensioned to produce axial
rather than radial alignment.

Dot
Dor

Sl Sr

G
Kul Kur

Kll Klr

Kv

Fig. 2.8.4The locator key positions


Figure 2.8.4 and adjustments.
The locator key positions and adjustments.

170
Steam Path Component Alignment and Stage Spatial Requirements

Movement of keys Kll and Klr achieves adjusting for the


lower half in radial alignment. Support is also provided by the verti-
cal key Kv, which is intended to assist in alignment and prevent
diaphragm sag at high temperatures under the effects of its own
weight. The key Kv also takes the torque reaction developed in the
stationary blades and so maintains transverse alignment. Triangula-
tion of these three keys is undertaken at assembly to allow close
clearance on final assembly.

The upper and lower halves can be assembled by two distinct


methods. First, the upper half can be located in the upper half-cas-
ing with the horizontal joints of the inner web and outer ring being
common or in line with the upper half- casing half-joint. When the
upper half-casing is lowered into position over the lower, the
diaphragm upper half locates on the lower. The joint keys are
engaged to provide both location and, to a certain extent, provide a
steam seal. An alternative design allows the upper half-diaphragm to
be bolted to the lower half using studs (shown at positions Sl and
Sr). The upper half-casing is then lowered over the assembled
diaphragm. A steam seal is formed between the casing and
diaphragm outer ring by the steam pressure acting across it. In each
case, the steam path discharge diameters (Dot and Dor) can be
achieved within design limits.

Inserted blade design. This design locates stationary elements


within a casing or blade carrier. Their diameter is adjusted relative to
the rotating elements to provide an acceptable steam path. The radi-
al location of these stationary blade elements, assuming the base
dimensions of the individual blades are correct, depends entirely
upon the shoulder diameter of the root groove in the casings or car-
riers, which is shown as Ds in Figure 2.8.2. The steam path dis-
charge diameters remain at their correct location relative to the rotat-
ing blades (assuming the casing remains concentric).

171
Turbine Steam Path Maintenance and RepairVolume One

Other stationary components. Within the steam path are a num-


ber of components (packing heads and seal components, in addition
to the stationary blade elements) required to guide and limit steam
leakage. Keeping foreign materials out is as important as keeping
steam in. Oil used in the bearings, for instance, is not allowed to
enter the steam path, particularly in the low-pressure regions where
sub-atmospheric pressure exists. Under certain circumstances, such
pressure could suck oil into the steam system along the central shaft.

True
Dor center
Adj/h Adj/h

Kll Klr

s Klr

Adj/v
Kv D C

Fig. 2.8.5Details of key support


Figureand adjustment.
2.8.5
Details of key support and adjustment.

The component adjustment process


When a lower half-casing is set in a horizontal position (relative
to the bearing pedestals and those bearings supporting the rotor
within that casing), it becomes necessary to adjust the various sta-
tionary portions within the casing that locate about the rotor to
ensure concentricity. Consider the diaphragm lower half shown as
Figure 2.8.5, which is held in position by keys Kll, Klr, and Kv.
Each of these keys has adjustment. Keys Kll and Klr can be
adjusted in the horizontal position by amounts Adj/h. They can

172
Steam Path Component Alignment and Stage Spatial Requirements

also have their vertical height adjusted by change of the shims S


(shown by detail in the inset to this figure). This vertical height is
adjusted until the half joints of the diaphragm (D) and casing (C)
are level. The vertical key (Kv) is also adjustable in the vertical
position by an amount Adj/v, which may require modifying, or
changing the key.

Methods of alignment, casing to casing


There are currently two methods available for the alignment of
the components to ensure their bore and (in the case of blading) their
discharge diameters Dor and Dot are at design values. These
methods are described below:

Piano wire method. This method requires a tight piano wire be


placed on the casing at the midpoint between the horizontal sur-
faces. This establishes the true center of the major component
the casing. The requirements of this method are that a wire of known
diameter and material properties be suspended between centers of
known length, with a weight suspended from one end.

Figure 2.8.6 shows a wire of known properties stretched by use


of a weight W between centers C-C over a length L. The wire
is tensioned and has uniform stress along its length. At the mid-span
position is a maximum defection (D.). At any other position (X),
the deflection (d) can be determined from the expression for a
catenary. More often a sag chart is available to determine the
diameter, mechanical properties, and weight used over a variety of
lengths (L.)

With knowledge of the wire sag at any axial location, the indi-
vidual stationary elements can be aligned within the casings. The
casings can be set relative to their neighbors, with allowance made
for the various bearing settings required to account for the various
phenomena (as previously discussed).

173
Turbine Steam Path Maintenance and RepairVolume One

C C
L
X s
P
D d
w

Fig. 2.8.6The Catenary Sag Figure 2.8.6


of a piano wire.
The Catenary Sag of a piano wire.

Laser optical alignment. A more recent development is the use


of laser optical alignment techniques. A beam is established between
two reference points and the various components are aligned to this
beam. Figure 2.8.7 shows the process of lining out a section, show-
ing a bracket being rotated to various positions around the inner sur-
face to be adjusted. Using readings from these various positions
adjustments are made to achieve concentricity.

Fig. 2.8.7Laser optical alignment.

174
Steam Path Component Alignment and Stage Spatial Requirements

Checking with leads


A means often used to check alignment and clearances is to
place lead strips at various axial positions along the section. The
rotor is then lowered into position, with the weight of the rotor
crushing the leads locally so their thickness occupies the radial
clearance at the top and bottom dead center. When the rotor is
removed, the thickness of the lead is measured, providing an accu-
rate reading of the cold setting unbolted clearance. This is also often
performed with the bolts at the horizontal joint, both bolted and
unbolted, to achieve true readings.

UNPREDICTABLE FACTORS
AFFECTING DESIGN
CLEARANCE
Thus far we have considered predictable factors that influence
the hot running axial and radial clearances within the steam path.
These are factors the design engineer examines during the various
stages of design, to the accuracy with which components can be set
at cold stationary conditions. The goal is to enable the unit to oper-
ate with a minimum of risk of running interference, and the per-
formance of each stage maximized (within the limitations set by
mechanical considerations).

There are, however, two groups of considerations influencing the


actual running clearances. The first have been considered; the sec-
ond group is those factors that are unpredictable. These phenomena
occur normally as a consequence of operating conditions, operating
excursions, or major transients of inlet steam parameters. There are,
however, other considerations to these unpredictable phenomena
that occur as a consequence of stress relief, extended exposure to the

175
Turbine Steam Path Maintenance and RepairVolume One

high temperatures and pressures of the steam, and possibly other fac-
tors of design or manufacture.

Unpredictable phenomena can modify the shape of the compo-


nents by various forms of permanent deformation, which as a con-
sequence affects the initially selected dimensional settings. What is
particularly difficult with these phenomena is that while the proba-
bility of their occurrence once or twice during the life of a unit is
almost certain, and some allowance can be made for them, their
extent is unpredictable. The severity of the deformation or the mod-
ifications that result is normally dependent upon the type and sever-
ity of the condition causing the modification.

Various phenomena can occur, but the most common form of


deformation or distortion follows:

Casing ovality (bowing)


Casing walls are not of uniform thickness at all tangential loca-
tions. There is a massive heat sink at the horizontal joint where the
thick flanges can cause temperature differentials under conditions of
loading and temperature transients. For these reasons various loca-
tions of the casing accept and reject heat to the steam at different
rates while operating at the same temperature and under different
operating conditions and situations. These temperature effects
together with the possibility of residual stressescan cause the tur-
bine casings in most instances to assume an oval form.

Figure 2.9.1 shows a schematic section through a turbine half-


casing. The casing has an internal diameter of Dx and a vertical
half height of Vx = Dx/2. This is the design specified requirement.
The halves are held in their operating position by a series of bolts at
the horizontal joint. The boltholes are also shown. Under the action
of high temperature steam, these dimensional conditions can

176
Steam Path Component Alignment and Stage Spatial Requirements

Vx

Ft

Dx

Fig. 2.9.1The design conditionsFigure


at a section
2.9.1through an upper half casing.
The design conditions at a section through an upper half casing.

change, and the design roundness can be lost. There are two condi-
tions that can occur:

A reduction of the horizontal joint diameter Dx to Dx -


2.dRx. Under these conditions, the vertical half height (Vx) will
increase by some amount (+dVx), as shown in Figure 2.9.2(b). This
condition can be damaging to the extent the casing will close down
on the rotor at the horizontal joint, and in the case of severe distor-
tion, can make it difficult to lift the top half. It is also possible the
horizontal joint studs will bind in their holes and become necessary
to be bored out rather than removed by normal means.

In some instances these thermal effects cause residual stresses to


be induced in the casing halves at a level that will allow the rotor to
be turned while the casing halves are coupled. However, as soon as
the studs are released, the casing moves in response to the residual
stresses and clamps on the rotor. These are situations that must be
evaluated in terms of removing the top half, and the potential this
poses for damage to the rotating blade rows.

An increase in the horizontal joint diameter Dx to Dx +


2.dRx. In this situation, the vertical half height (Vx) will decrease
by some amount (-dVx), as shown in Figure 2.9.2(a).

177
Turbine Steam Path Maintenance and RepairVolume One

-dVx

Vx

Ft

+dRx Dx +dRx

+dVx

Vx

Ft

Dx -dRx
-dRx

Fig. 2.9.2Casing ovality. InFigure(a) the 2.9.2


casing has an increased horizon-
talCasing
joint width, and
ovality. In in
(a)(b)
thethe horizontal
casing has anjoint is reduced.
increased horizontal joint
width, and in (b) the horizontal joint is reduced.

This again is a situation where the casing can clamp the rotor,
either at low rotating speeds, at standstill, or when the studs are
released. However, this clamping is in the vertical direction. While
this can cause rotating blade damage by bending themparticularly
as the unit comes to restand the deformation becomes a maxi-
mumthis form of deformation will not cause a horizontal clamping
of the rotor and prevent disassembly. However, the studs may need to
be bored out if the horizontal movement of the casing is sufficient to
bend them in the casing halves on release of the tightened studs.

If either of these deformation movements of the casing occur, it


is possible for the radial sealing arrangements to be damaged if the
movement is large and occurs while the rotor is turning at any speed.

178
Steam Path Component Alignment and Stage Spatial Requirements

Figure 2.9.3 shows grooves cut into the casing at the horizontal joint
by the rotating blade seal as the casing has moved radially (in Fig.
2.9.2(b) condition). This situation could become sufficiently severe
that blades could be bent and the rotor grabbed. This will have the
effect of destroying the sealing capability at the rubbed location, and
depending upon the method of seal construction, can have signifi-
cant long-term effects on the efficiency of the unit, if they cannot be
replaced.

Fig. 2.9.3Casing damage produced by heavy rubs of seal strips carried on the rotating
blades. This introduces grooves on the casing at its horizontal joint.

With the condition shown in Figure 2.9.2(b), it is possible ovali-


ty can worsen when the horizontal joint bolts/studs are released. At
that time, the rotor will move in at the horizontal joint and two sit-
uations can arise:

Threaded components at the horizontal joint bind in the cas-


ing holes, making it difficult to lift the top half casing. This
normally requires the bolts/studs be drilled out

179
Turbine Steam Path Maintenance and RepairVolume One

The rotor is grabbed by the inner surfaces of the casing. This


would make it difficult to lift the blades that are in tight con-
tact with the casing by bending, without damage

Ovality can be corrected (to the extent the unit can be reassem-
bled and returned to service) by the controlled application of heat.
The means to make such corrections are discussed in greater detail
in chapter 7.

Casing humping (hogging)


A second distortion mode that can effect a casing is hump-
inga vertical displacement upwards from the original flat posi-
tion. This can result as a consequence of certain operating phenom-
ena (from rubs to residual stresses), possibly induced by excessive
positive or negative temperature ramps. However, the most common
cause is the result of the rotor being brought to rest and left station-
ary while hot gases (steam or air) remain in the casing. Under these
circumstances, hot gases in the casing rise and overheat the upper
half, which expands relative to the lower. Upon cooling, the casing
horizontal joint will not reassume a totally flat position. The effect of
this on steam path alignment is to make the rows of stationary blades
rise and become misaligned relative to the rotor. Affected blades are
located either in the casing, a blade carrier, or diaphragmparticu-
larly near the center span. Methods of correcting this condition are
discussed in chapter 7.

Horizontal joint leakage


As a consequence of operational distortion of the casing as dis-
cussed, there is always the possibility that leakage will occur
between the surfaces of the horizontal joint. This situation must be
corrected, to prevent steam bypassing the blade rows and lowering

180
Steam Path Component Alignment and Stage Spatial Requirements

efficiency, and to prevent steam reentering the steam path, possibly


causing a two/rev stimulus.

The most severe form of damage that can result to the horizontal
joint surfaces is when water is present in the steam and a washing-
erosion material loss occurs. In this situation, it is normally necessary
to make a weld repair. If the opening has been present for some time,
it is possible there could be deposits on the surfaces (Fig. 2.9.4).
These deposits can in fact help seal the gap and help reduce the
leakage quantity, but they do not correct the situation. Figure 2.9.5
shows the leakage pattern of superheated steam leaking past a sta-
tionary blade row. There are also deposits present on this surface.

Fig. 2.9.4The leakage path between the horizontal joint of a high pressure casing.
Had the steam contained water there would have been washing erosion at these loca-
tions.

181
Turbine Steam Path Maintenance and RepairVolume One

Fig. 2.9.5Surface deposits on a horizontal joint surface where there has been an
opening for an extensive period, and the unit has operated with high carryover from the
boiler.

Casing boring after adjustment


Even after heating corrections, many of the damaging conditions
that occur in the casings will retain a degree of out-of-roundness
in the shells. While much of this may be acceptable, it may some-
times be necessary to undertake corrective machining and modify
the form of the components located in the shell. Such decisions to
machine must be made only after careful evaluation of the total
requirements. Figure 2.9.6 shows a boring bar mounted to a high-
pressure inner casing. In this case, the bar is machining the cylindri-
cal faces to return the section to a fully concentric condition.

182
Steam Path Component Alignment and Stage Spatial Requirements

Fig. 2.9.6A boring bar being used to machining a casing and achieve local concen-
tricity.

Diaphragm dishing
This condition occurs when the diaphragms assume a degree of
dishingan axial deflection of the vanes and inner web in the direc-
tion of the pressure gradient. This effect is generally the result of either:

excessive pressure at high temperatures, causing creep defor-


mation (as discussed previously and shown in Fig. 2.6.1,
with a total deformation of dxd). This deformation may not
be to the same axial extent at all tangential positions

the effects of a hard axial rub between stationary and rotat-


ing portions of the unit while it is in operation

183
Turbine Steam Path Maintenance and RepairVolume One

The net effect of this deflection is for the diaphragm to deflect


axially downstream (in the direction of the pressure gradient). This
effect consumes the axial clearance between the diaphragm face and
the following rotating row wheel.

Dishing can also modify the sealing clearance and tooth setting
between the rotor and diaphragm inner seal diameter. If hi-lo teeth
are used, this can cause the seal strips to rub (i.e., it will modify the
tooth settings). This is discussed in chapter 7, and is illustrated as
dimensions g and h in Figure 2.12.25.

In fact, all of these effects are discussed in chapter 7, and Figure


7.7.2 shows deposits from a rotating wheel on a diaphragm inner web
after a sustained heavy rub. This diaphragm had been hardened to
about BHN 380 in the region of the deposited material. This
diaphragm deflected downstream and also took an elliptical form [see
the following section, Axial deformation of stationary (body) blades].

In some earlier designs (older units still in service), the selection


of the weld prep, filler materials, and welding process (fabrication
sequence) has had an effect on local strength. If uneven around the
half ring, it allows a different degree of axial deformation, which
contributes to an uneven dishing effect.

Figure 2.9.7 shows the measured values of axial deformation in


a diaphragm half, which has operated for 150,000 hours at high tem-
peratures. The deformation (dxd) was measured on the hook face
of the diaphragm seal-locating slot (which was the cleanest surface
available) and had been made free of deposits. Note that the defor-
mation is greater towards the horizontal joint. This form of deforma-
tion can be corrected by heat, with axial forces applied to return the
element to its original position. It may be necessary to undertake
some skim cutting after heat treatment, particularly on the seal sur-
face on the packing ring.

184
Steam Path Component Alignment and Stage Spatial Requirements

Note: While dishing from creep deformation can be corrected by


various means, creep is a metal failure at the grain boundaries, and
correcting this condition will not reestablish the grain boundary
strength. Such correction is therefore considered temporary.

Radial
Location for
measurements 22.5
of deformation degrees

5 6
4
3 7

2 8

1 9
Steam pressure
side

dxd
deformation "dxd"

0.10
0.08
Axial creep

0.06
0.04
0.02
0
1 2 3 4 5 6 7 8 9
Tangential Position

Fig. 2.9.7The measured values of axial deformation


Figure 2.9.7 dxd at the seal slot on a
diaphragm which
The measured has considerable
values exposure
of axial deformation to high
dxd at thetemperature
seal slot onand pressure differ-
a diaphragm which
entials.has considerable exposure to high temperature and pressure differentials.

185
Turbine Steam Path Maintenance and RepairVolume One

Axial deformation of stationary (body) blades


In the event stationary blades inserted in a casing are deformed
to the extent that rubbing contact could occur, it is normally far
more cost effective to replace the blades rather than attempt to
straighten them.

Outer Ring
(a)

Inner
Stationary Web
Blade
Row

+do +do
Do

Outer Ring
(b)

Inner
Stationary Web
Blade
Row

-do -do
Do

Fig. 2.9.8In (a) is shown a diaphragm,


Figure 2.9.8 in which the horizon-
tal
In joint
(a) ishas opened
shown by an amount
a diaphragm, +do
in which theon both sides.
horizontal Inhas
joint b)
the horizontal
opened by an joint has+do
amount closed
onby a similar
both amount
sides. In (b) the-do on
horizon-
both sides.has closed by a similar amount -do on both sides.
tal joint

Diaphragm concentricity
Just as the casings horizontal joint diameter will change under
certain load conditions, pressure, and temperature, so the diaphragm
can creep in or out at the horizontal joint. If this deformation

186
Steam Path Component Alignment and Stage Spatial Requirements

occurs, it will carry the stationary blade row to other than the orig-
inal diameters and cause the steam path to loose concentricity.
Under extreme conditions it can also produce rubs of the seal strips
onto the rotor body, coverband, or other location.

This type of distortionboth closure and opening of the hori-


zontal jointcan be corrected by jacking the joint to the correct
diameter (Fig. 2.9.8). The halves are then subjected to a stress relief
cycle, normally in a vacuum furnace. This should be done after the
steam path, the steam seal surface, and the horizontal joint have
been protected by a suitable compound (one that is inert and will
prevent oxidation). It may be necessary after this treatment to skim-
cut the steam seal face and horizontal joint to achieve an acceptable
seal. Alternately, it might be necessary to make a surface weld
deposit to build up the joint material. This will require machining
and (possibly) stress relief. If the joint has opened, it can be closed
mechanically using an external jack and then stress relieved as
described above.

Fig. 2.9.9The horizontal joint of a diaphragm which has deformed and moved radially
in. The horizontal joint key can be seen.

187
Turbine Steam Path Maintenance and RepairVolume One

Figure 2.9.9 shows the diaphragm horizontal joint gap produced


by two halves, after they have been removed from the unit and
clamped together. It shows the effect of a diametral change that has
occurred. In this stage, the horizontal joint has moved in, causing the
horizontal joint to open at a position closest to the rotor. The key can
be seen in the gap now existing between the top and bottom halves.

Rotor bends
Phenomena associated with the operation of turbines can induce
bends into the rotor. Normally, if these are not too severein excess
of about 0.006-0.009"the rotors can be balanced and can contin-
ue to operate. However, with this form of operation, the rotating
blade rows at and opposite the position of the bend maximum con-
dition are no longer concentric with the stationary row, and the lap
will be consumed at both the inner and outer diametral positions.

Operating with a bent rotoreven if adequately balanceddoes


not represent a satisfactory condition and should be corrected as
soon as possible. However, it may be necessary to accept such oper-
ation for a period in order to return the unit to service.

STEAM PATH
AREA REQUIREMENTS
In assessing dimensional and tolerance requirements of the
steam path, it is necessary to consider the objectives of the blade sys-
tem, relate these objectives to principal dimensions, and then con-
sider their possible influence on various performance criteria. This
process establishes blade geometry and certain radial alignment
requirements, assists in the selection of various elements, and pro-

188
Steam Path Component Alignment and Stage Spatial Requirements

vides guidance in selecting arrangements suited to achieving these


concurrent objectives.

In designing systems to achieve these two objectives, sufficient


information must be generated concerning mechanical loads and
thermal conditions within the steam path, so component selection is
facilitated. Other concerns such as stress levels within the local envi-
ronment and frequency effects, can also be investigated. After com-
ponents have been selected and sized, manufacturing techniques are
chosen and tolerances applied to ensure a satisfactory product.

For the purpose of this section, the information that the design
process should generate and specify (and the owner should be famil-
iar with) concerns the relationships existing at various locations and
areas, and the rates at which it varies throughout the steam path. In
defining the form the various components take, the designer selects
or defines tolerances within which the individual components must
be produced, how they are assembled, and the tolerances they pro-
vide to ensure area requirements are met.

The area relationships within the steam path that should be con-
sidered are:

The area of the passage formed between stationary and rotat-


ing blade profiles, and its variation from inlet to discharge

The discharge area of individual passages and its variation as


a function of blade height

The variation of discharge area from passage to passage with-


in the row.

The total stage discharge area

The leakage area occurring at seal points between stationary


and rotating portions of the unit

189
Turbine Steam Path Maintenance and RepairVolume One

As stated, any units performance will be influenced by the


design of these areas and the manufacturers ability to consistently
meet them within design specified tolerances. Normally, after steam
path components are manufactured and assembled, the manufactur-
er will undertake a series of audits to determine whether the overall
dimensional tolerances (both of the individual components and their
assembled arrangement) are within acceptable limits.

When the unit is removed from service, damage is found, and


repairs are undertaken. It is necessary to perform this work in such a
manner that the mechanical integrity of the components is retained
(or improved) and original area relationships are preserved. Although
there is some latitude in these values, if it is known a repair proce-
dure will cause a significant variation from the design-determined
values, then a careful evaluation of the possible effects on the per-
formance level should be made.

THE STAGE OPERATING


DEFINITION
An important aspect of the design phase has the engineer estab-
lishing the dimensional requirements of the various components
comprising the steam path, the tolerances within which they are to
be produced, and the material to be used. In addition to these obvi-
ous requirements, surface finish must be defined, special process
procedures possibly identified, and the total quality (conformance
level) to be achieved.

During original component manufacturewhen a component


outside normal tolerances is produced, or when a nonconforming
situation occurs during assemblyit is the design engineers respon-

190
Steam Path Component Alignment and Stage Spatial Requirements

sibility to evaluate the condition and provide a disposition that will


help ensure the unit meets its expected performance levels.

A nonconforming condition is one in which the existing ele-


ments no longer meet design requirements, i.e., individual compo-
nents have deteriorated in their mechanical properties or surface and
dimensional requirements. They have probably lost surface material,
and no longer comply with the requirements established by design.

If deterioration is found at a maintenance outage, it is again the


responsibility of a design or plant engineer to rule on the condition
and develop an acceptable disposition. However, in this case, the
disposition may not return the unit to the same condition it pos-
sessed as a new addition to the system. While such a disposition may
return the unit to an acceptable conditionperhaps even possibly
long-term reliabilityit may also compromise efficiency. Often this
has to be the case until a final dispositionpossibly involving the
provision of replacement partscan be implemented.

Subsequent evaluation of these operational nonconformances


may reveal two (or more) possible scenarios for correcting the situa-
tion and returning the unit to an acceptable level of performance.
Some of these options may depend upon there being available suit-
able techniques for either repairing or refurbishing the condition,
others may require the use of replacements parts. It is advisable for
the maintenance engineer to remain aware of these techniques, the
limits of their capability, and how (and if) they could be applied in
the situations being considered.

The dimensional requirements discussed in the following sec-


tions represent conditions to which the components were originally
manufactured, and should preferably be returned after they have
been rebuilt and/or adjusted during the maintenance repair process.
In the design-specified condition, elements are arranged to ensure
steam expends its thermal potential energy and is converted to rota-
tional kinetic energy within the rotor. The stages are also arranged so

191
Turbine Steam Path Maintenance and RepairVolume One

the steam flows through them with a minimum of loss, and the effi-
ciency of energy conversion is maximized.

During the original manufacture of turbine steam path compo-


nents (and during their repair and replacement), certain spatial rela-
tionships in the axial, tangential, and radial directions must be met,
while others (by preference) should be met to assist in optimizing
unit performance.

Purchasers of steam turbine equipment must ensure that suppli-


ers adequately define manufacturing needs, and have an acceptable
and controlled system for ensuring these are consistently met. The
purchaser must also determine that a system exists for evaluating and
disposing of nonconforming conditions at a level (within the manu-
facturers engineering organization) competent to make such deci-
sions.

Engineering tolerances are selected by design to help ensure per-


formance targets are met without incurring costs that would make the
equipment too expensive to purchasers. Where appropriate, compo-
nents should be able to be disassembled for repair or replacement.
Tolerances should also ensure components can be interchanged,
both within similar units in a station and between stations.

STEAM PATH
COMPONENT ARRANGEMENT
(AXIAL/RADIAL DIRECTION)
Irrespective of the degree of refinement used to manufacture and
assemble steam path components, these parts must be correctly
aligned within the unit to permit the controlled expansion of the
steam from one row of elements to the next. If this is not the case,

192
Steam Path Component Alignment and Stage Spatial Requirements

then neither sophistication of design, nor quality of manufacture will


produce a unit with an acceptable level of performance.

Consider the cross section of the two stages shown in Figure


2.12.1. An impulse stage is shown in (a) and a reaction stage in (b).
These figures show the basic spatial requirements for clearance,
lap, and blade setting in a plane through the stage.

Ca
Cro

Lno
Lbo
Dot Det *Dit
Co *Dot
Dit
Lni

Lbi

*Dir *Dor
Dir

Cri *Cri

Dor

Figure 2.12.1(a)The spatial requirements of an impulse stage in the axial/radial


Figure 2.12.1(a)
direction.
The spatial requirements of an impulse stage in the axial/radial direction.

Obviously, these alignment and adjustment settings of the steam


path components must occur in the cold stationary condition. Any
relative movement of the parts that occurs during operation must be

193
Turbine Steam Path Maintenance and RepairVolume One

determined, and allowance for them made in establishing these cold


settings. The principal factors influencing the hot running condi-
tionand therefore the cold condition alignmentwere discussed
earlier.

Caulking

Cro

Det
Lbo
Dnt
Coo
Dit Cao
Dot

Cai
Coi

Lbi

Cri Dnr
Dor
Dir
Der

Fig. 2.12.1(b)The spatial requirements of a reaction stage in the axial/radi-


Figure 2.12.1(b)
al direction.
The spatial requirements of a reaction stage in the axial/radial direction.

The design process accounts for (and evaluates) the influence of


the various operating phenomena and determines certain critical
alignment or setting values in the components, as well as their
required position relative to each other. Establishing cold settings
ensures the clearances C and lap L dimensions are consistent with
design requirements, and as predicted to occur during operation.

194
Steam Path Component Alignment and Stage Spatial Requirements

To assist (and promote) steam flow from one row to the next with
a minimum of disruption, the inlet height of any row receiving steam
is chosen to have a lap L. This makes the inlet height of a station-
ary or rotating blade row marginally larger than the discharge height
of the row from which it is receiving the steam. Because of differ-
ences in axial gap between stagesand to allow for the radial flow
effect of the steam at exit from any stagethe value of lap will vary
from row to row. The radial flow effect will require the outer diame-
ter laps to be larger than the inner laps. However, if the outer lap
becomes too large, there is a possibility of flow separation and
excessive turbulence being induced at the entry to some rows.

Clearance requirements must be considered in both the axial


and radial direction, depending upon stage location and arrange-
ment. Critical clearances may, in addition, occur at the inner and
outer extremities of the stage. The leakage areas occurring at any seal
point provide an opportunity for steam to bypass the blade elements.
This leakage constitutes a direct loss to unit output and efficiency.
Efforts made to control lap and clearances at the design values can
therefore represent considerable gains in maintaining unit efficiency.

The various components of the steam path must be machined


and assembled so that during hot running conditions the various por-
tions of each stage are in an acceptable alignment during normal
and transient operation. If this is not done, then stage efficiency will
be reduced because steam will not flow from one element to the
next without incurring higher losses. Should rubs occur, they could
lead to the opening of clearances. Under the most severe conditions,
such rubs can cause burning of the surfaces in contact, which will
affect material properties and could eventually cause failure of the
affected parts.

During normal operation, the steam path parts must be so aligned


that steam will flow from one row to the next and incur minimum
aerodynamic losses due to steam path irregularities or misalignment.

195
Turbine Steam Path Maintenance and RepairVolume One

Steam path efficiency is not an important consideration during tran-


sient conditions of load, temperature, or other cycle conditions that
affect either clearances or lap. However, the clearances and lap for
normal operation must be chosen so rubs and interference do not
occur during these transients. This requirement will often mean the
design process will specify values of clearance and lap to accommo-
date transient requirementsselected values that are larger than opti-
mum for normal operating conditions. Although this does lead to rel-
ative inefficiencies of the steam path, such conditions cannot be
avoided. To ensure acceptable alignment during operation, it is nec-
essary to predict the operating conditions using them to calculate the
cold setting requirements that will provide for the four major pre-
dictable influences of movement, and also make adequate allowance
for the potential affect of transient operation.

To achieve acceptable cold alignment, three major areas (or


regions of tolerances) must be controlled during the manufacturing
and assembly phases. These areas must also be reviewed for com-
pliance at completion of any repair/refurbishment actions, if these
repair techniques are capable of affecting the basic dimensions.

These dimensional requirements represent conditions that must


be correct in the individual components and assemblies before erec-
tion begins, or it will not be possible to achieve an acceptable unit
alignment. These areas are:

rotor machining and assembly

diaphragm and stationary blade manufacture

seal production and adjustment

The quality of the steam path is established before the unit is


delivered to the site and field-erected. It is established by the design
process, the method by which the design requirements are met, and
finally, compliance with the three groups of considerations listed
above. It is necessary to know the requirements and compliances

196
Steam Path Component Alignment and Stage Spatial Requirements

needed to design these components so they can be confidently


installed in the unit.

Rotor machining and assembly requirements


(axial direction)
The rotor is an assembly on the main central spindle of rotating
blades, coverbands, and tie (lacing) wires. It can also include
shrunk-on wheels, coupling flanges, seal rings, and other elements.
The rotor can also be that portion of the unit on which the thrust col-
lar is located, and therefore the major component responsible for the
total axial position of the unit during operation.

A typical high- or intermediate-pressure, double-flow rotor is


shown in Figure 2.12.2. It indicates the major axial dimensions that
must be achieved before the rotor is delivered for assembly.

O m4
m3
m2
t u
m1
u t
t

n
n

u
f

Fig. 2.12.2The basic control dimensions


Figureon2.12.2
a rotor. In this example a double flow
element, without a thrust
The basic block.
control dimensions on a rotor. In this example a double
flow element, without a thrust block.

197
Turbine Steam Path Maintenance and RepairVolume One

Certain dimensional requirements must be met to ensure the


rotor can be assembled into the casing and fit among the stationary
blade rows to achieve the clearances required by design. In defining
these dimensions, a reference point must be established. If the rotor
carries the thrust collar, then it is selected as the reference point. If
(as shown in Fig. 2.12.2) there is no thrust collar, a coupling face is
selected. Such a face is shown as position O-O in Figure 2.12.2
(an alternate location would be one close to the center of the rotor
span). A coupling flange can be adjusted at erection by using spac-
er plates of various thicknesses between the flanges. The important
requirement in Figure 2.12.2 is the relationship of the various dimen-
sions to each other. When a thrust collar is used as reference, the
relationship of these dimensions to each other is important, but since
no site adjustment is possible, the absolute dimensions to the thrust
face are also critical. Ultimately, all axial dimensions are checked
and set relative to this thrust collar, whichever rotor it has been pro-
duced on

Consider the dimensions m in Figure 2.12.2. They reference


the position of the individual stages to the reference face. Integral
and shrunk-on wheels must be accurately spaced relative to the cou-
pling flange and each other to permit the diaphragms or stationary
blades to be assembled between them and maintain design clear-
ances [Fig. 2.12.1(a) and (b)].

In Figure 2.12.2, the important wheel dimensions are shown as


m1-m4. This establishes the wheel pitching along the axial length
of the rotor. These dimensions, combined with the wheel thickness
t, establish the axial gap available for the stationary blade row. This
final gap between the wheels is ut. The stationary blade row must
be assembled into this gap to provide clearances Ca and Co [as
shown in Figs. 2.12.1(a) and (b)] and be suitable for hot and cold
rotor conditions.

Note: When considering the Ca and Co clearances, it is


important in those stages with an overhung coverband (Fig. 2.12.3

198
Steam Path Component Alignment and Stage Spatial Requirements

and 2.12.4) to normally define clearance as shown in these figures


(i.e., the clearance from the stationary blade rows to the knife edge
of the seal). This clearance is set (the coverband is machined) after
the rotor is assembled. It is also important to ensure the axial clear-
ance to the blade inlet edge does not compromise the running clear-
ance under the effect of transients.

Seal

m o
Cr

Fig. 2.12.3An angled


Figuretip2.12.3
seal system employ-
ing An
oneangled
axial and one system
tip seal radial strip.
employing one
axial and one radial strip.

Cold setting
position

Ca Cr

Fig. 2.12.4A tip sealing


Figure arrangement
2.12.4 for a coni-
cal tipsealing
A tip section.
arrangement for a conical tip section.

With the considerations mentioned above, the blades must be


assembled to the wheel and two major dimensional requirements

199
Turbine Steam Path Maintenance and RepairVolume One

must be controlled. First, the coverband must be spaced in its axial


extremities and set so interference does not occur. If the band con-
tains an axial seal or provision for a portion of a radial seal platform,
these must be spaced correctly (Fig. 2.12.3 is an axial seal cover-
band.) Here the band has an overhang portion o and a required
knife-edge thickness x. These dimensions are normally achieved by
trimming after the band is assembled to the blade tips. If the band
also contains a platform (Fig. 2.12.4), then the platform must be axi-
ally spaced so the seal strip above it during normal operation is in
the correct position.

Figure 2.12.5 shows the root platform a-a which must be accu-
rately placed. This is of considerable importance if a seal is produced
in the root area, and requires a platform that is machined integral
with the root. Such a radial seal is often produced on a stage with
tangential entry roots, and is produced by machining excess stock
left on the root form after the blades are assembled to the rotor.

Ca

Moving Blade
Fixed Row
Blade
Row

a a

Cr

Fig. 2.12.5A rootFigure


sealing2.12.5
system used to minimize
leakage
A root to the inner
sealing sealused
system diameter.
to minimize leakage
to the inner seal diameter.

200
Steam Path Component Alignment and Stage Spatial Requirements

Some blade designsparticularly those required for close axial


clearances in a reaction type stageare produced and assembled to
the wheel with excess stock or material remaining on them for final
trimming. This is shown in Figure 2.12.6, where the stock is shown
as an axial amount s. This material is trimmed after assembly to
provide the correct axial spacing (m). Care must be exercised in
trimming this stock to ensure there is sufficient material below the
vane to support it. The material should not be removed to the extent
k exceeds the value shown as k2. If the machining compromis-
es the blade vane support, there could be an error in some axial
dimension, possibly relating to the root groove position.

k1
k2
k3
s

Fig. 2.12.6Material trimmed


Figurefrom the rot platform of a
2.12.6
rotating bladetrimmed
Material installedfrom
on athe
drum
rootrotor.
platform of a rotating
blade installed on a drum rotor.

Impulse rotor blades are located at the outer diameter of wheels,


which in high and intermediate pressure sections are formed integral
with the rotor central spindle. However, drum type rotors, because of
the greater number of stages required to expand the steam efficiently,
have their rotating blades located in grooves machined into the outer
radius of the rotor center drum section (Fig. 2.12.7). In machining

201
Turbine Steam Path Maintenance and RepairVolume One

these grooves it is necessary to locate the groove at position f and


also to control the neck width g. This ensures the axial spacing can
be achieved and there is no opportunity for the root block to twist in
the groove and modify the vane-setting angle.

f
g g

Fig. 2.12.7Closely spaced


Figurerotating
2.12.7blade rows mounted
on a drum rotor.
Closely spaced rotating blade rows mounted on
a drum rotor.

In establishing the correct axial position has been achieved for


these drum-carried blades, it is necessary to gauge that dimension
f has been achieved. This must be done by measuring from the ref-
erence face to the blade groove side (i.e., to f+g/2 or f-g/2).

On many rotors there is provision for machining the seals region


to ensure the steam sealing strips are correctly located relative to the
stationary portions that normally carry the seal strips. Figure 2.12.8
shows the region of castellations required for hi-lo strips. For the
seals to be effective, and to operate without interference during tran-
sient conditions, the shoulder location n must be correct, and then
the individual pitch (p) and castellation location (w) must be cor-
rect relative to the shoulder position (n). This is of considerable
importance on systems that employ the hi-lo tooth configuration,
where teeth are alternately located on the upper and lower castella-
tions of the rotating parts.

202
Steam Path Component Alignment and Stage Spatial Requirements

Wn

W3
W2

W1 p p

Figure
Fig. 2.12.8The control dimensions at a 2.12.8
castellated rotor location showing
The control dimensions at a castellated rotor location showing
critical dimensions.
critical dimensions.

Some designs utilize seal strips that are caulked directly into the
rotating portions of the unit. This is shown as Figure 2.12.9(a), where
the pitching (P) must be maintained at the design values located at
a known distance (j) from the reference face. The seals must be
machined to a maximum diameter Dk to form the seal with radial
clearance Cr. Designs are also used where seal strips are caulked
into both the rotating and stationary portion [Fig. 2.12.9(b)]. Here,
the pitching is the same, and the clearance between the rotating and
stationary seals is selected so rubbing interference will not occur
during transient conditions. The teeth are arranged to have seal
diameters Dko and Dki. The radial clearance is set at Cr at the
inner and outer seal positions.

Dk P P P

Fig. 2.12.9(a)Seal stripsFigure


mounted in a stationary
2.12.9 (a) portion of a unit,
showing Seal
control dimensions.
strips mounted in a stationary portion of a unit,
showing control dimensions

203
Turbine Steam Path Maintenance and RepairVolume One

t
P
p1
p2

Cr Cr

Dki
Dko P P P

Fig. 2.12.9(b)Inserted teeth mounted alternately


Figure 2.12.9 (b)in the stationary and rotating
portions. Inserted teeth mounted alternately in the stationary and rotating portions.

Rotor machining and assembly requirements


(radial direction)
In addition to the axial requirements, the rotor also must be radi-
ally in the correct position. The more important diameters are shown
in Figure 2.12.1(a) and 2.12.1(b). This requires that diameters must
be set to achieve lap L and clearances Cr consistent with design
specifications.

The considerations concerning radial clearances at many posi-


tions are shown in Figures 2.12.3, 2.12.4, and 2.12.5 as steam trans-
fers from blade row to blade row, and in Figures 2.12.9(a) and (b) at
shaft-end positions. The sealing requirements at diaphragm-to-shaft
positions, and under stationary blade rows, are also essential to min-
imize leakage losses and maintain efficiency at the highest possible
levels.

Consider the diaphragm shown in Figure 2.12.1(a). Here, the Di


and Do diameters define the steam paths at inlet to and discharge
from the stationary blade elements. These set the steam path height at
inlet and discharge to the blade row for the stationary portion of this

204
Steam Path Component Alignment and Stage Spatial Requirements

stage. Therefore, to achieve a suitable lap (Lbi and Lbo at inlet to


the rotating blade row), the rotating blades must be located with inlet
diameters Dit and Dir (Fig. 2.12.10). Similarly, at discharge from
the rotating blade diameters, settings must be at Dot and Dor from
the row to achieve the correct discharge area, and then the following
stationary blade row geometry arranged to achieve laps Lni and
Lno.

x Dot
Dit

Dm
outlet

Dm
inlet

Dir Dor

Fig. 2.12.10The control


Figuredimensions
2.12.10 for a blade at its
inletThe
andcontrol
discharge diameters.
dimensions for a blade at its inlet and

For the rotating blade row to achieve the design lap in the cold
stationary condition, it is necessary to achieve the inlet and dis-
charge diameters as shown in Figure 2.12.1(a) and (b). These diame-
ters are achieved by mounting the blade to the wheel or rotor and
then, where necessary, using caulking to achieve a final diameter fix
(Fig. 2.12.11). Other diametersparticularly those associated with
producing platforms to form radial sealsare normally machined
after assembly by the removal of stock material left on for that pur-
pose. Such radial adjustment is also achieved in certain designs by
rolling side grips to move the blade radially outwards under the
action of the rolling force.

205
Turbine Steam Path Maintenance and RepairVolume One

Packing
Piece

Fig. 2.12.11A packing piece


Figure wedged between the
2.12.11
blade root andpiece
A packing wheel rim. between the blade root
wedged
and wheel rim.

Discharge annulus form


Further diametrical requirements of the blade rows (stationary
and rotating) are to ensure that the variation of radial height and
placement of blade to adjacent blade is maintained at design toler-
ances throughout the complete row. Figure 2.12.12 shows the dis-
charge portion of adjacent passages in a blade row, either stationary
or rotating. These passages have a discharge height h and inner
and outer design diameters of Di and Do, respectively. This fig-
ure shows the tolerance bands dDi and dDo within which the
inner and outer sidewalls must be located to ensure a smooth tran-
sition of steam flowing from one row to the following.

These tolerances may also be stipulated by the manufacturer as


the permissible radial variation for the passage position (dRo and
dRi) from the design position. In addition, tolerances can be
expressed for a radial variation (dSi and dSo) from one passage
to its neighbor. These dS passage variation values are not neces-
sarily the same as the diametral tolerance bands, which apply to the
total diametral variation within a row. The total requirement may also
be expressed as a value dh, within the diametral tolerances dDi
and dDo. Height tolerances refer to the total variation of the pas-
sage height, while diametral and radial tolerances refer to the pas-

206
Steam Path Component Alignment and Stage Spatial Requirements

sage position. In establishing permissible height variations, the


design engineer must consider the effect such variations will have on
the total discharge area, and set these height tolerances accordingly.

-dRo +dRo
dDo dSo

Do
dDi dSi
h

+dRi -dRi

Di
Stage Dimensions, (Design).
h = Blade Radial Height.
Do = Outer Diameter. 'dDo' = (+dRo)+(-dRo).
Di = Inner Diameter. 'dDi' = (+dRi )+(-dRi ).

Stage Tolerances.
dDo = Total tolerance on the outer diameter.
dDi = Total " " " inner "
dSo = Total stage tolerance at outer diameter.
dSi = Total " " " inner "
+dRo = Plus radial tolerance at outer diameter.
-dRo = Minus " " " " "
+dRi = Plus " " " inner "
-dRi = Minus " " " " "
Fig. 2.12.12The control dimensions and definition of the toler-
Figure 2.12.12
ances around a blade row discharge.
The control dimensions and definition of the tolerances
around a blade row discharge.

When any form of refurbishment is undertaken on either station-


ary or rotating blade rows, it is necessary to take no actions that will
compromise area tolerances, or at least minimize those that may do
so. In many older designs using cast iron or steel diaphragms, there
can be a considerable radial variation of the inner and outer sidewall
positions. It is also common to see these diameters affected by solid

207
Turbine Steam Path Maintenance and RepairVolume One

particle and water erosion. During maintenance outages, when these


elements are being refurbished, it is often possible to take some
remedial action that will reduce the total variation dsi and dSo
within a row. Figure 2.12.13 shows a finishing process on a weld-
fabricated diaphragm, where the outer sidewall is dressed to achieve
the correct diametral dimensions.

Fig. 2.12.13Finishing the outer side wall of a weld constructed diaphragm.

Many units, when removed from service, will be found to have


lost material from their sidewalls. This loss most commonly occurs on
the outer wall, and is aggravated by the radial flow effects on the
steam and the eroding material that it is transporting. This condition
is often seen in a last-stage cast iron diaphragm outer wall [Fig. 1.9.20
of chapter 1, Fig. 2.12.14(a)] where moisture impact and washing
erosion has removed material. In Figure 2.12.14(b), material has been

208
Steam Path Component Alignment and Stage Spatial Requirements

Fig. 2.12.14(a)The outer side wall of a cast iron diaphragm, showing material loss
due to water washing and impact erosion.

Fig. 2.12.14(b)The outer sidewall of a high pressure fixed blade row, where material
has been lost due to solid particle erosion.

209
Turbine Steam Path Maintenance and RepairVolume One

lost due to solid-particle erosion of an outer sidewall. These are con-


ditions that must be assessed in terms of their potential effect on stage
and unit performance, and decisions taken concerning the need for
corrective action.

B
A

dR

Radial
mismatch
dR
dR
A B

dT

Fig. 2.12.15Outer sidewall


Figure discontinuity and
2.12.15
shaping to correct
Outer sidewallthe passage shape.
discontinuity and shaping to
correct the passage shape.

When the abutment of individual stationary or rotating blade ele-


ments forms the passage, it can cause a discontinuity in the mainte-
nance of discharge diameters and the form of the expansion passage.
Such a condition occurs at one diametral locationeither inner, or
(as shown in Fig. 2.12.15) the outer. The effect of this discontinuity
is to introduce turbulence in the expansion passage, causing vortices
and eddies to be carried into the following row of stationary or rotat-
ing blades. This carryover has the potential to induce aerodynamic
losses and introduce stimuli that can create vibration in the blade
row. In terms of its effect on efficiency, this is more significant in
shorter blade rows, where the radial disturbance represents a greater
percentage of the total radial height. As shown in Figure 2.12.15, a

210
Steam Path Component Alignment and Stage Spatial Requirements

mismatch such as dR should be blended to minimize the carry-


over. However, the effect of the diameter change must still be with-
in the tolerances for diameter, as defined by Figure 2.12.12.

There may also be diametral discontinuities at the inner diame-


ters of rotating blades when it is necessary for root blending radii to
meet at the root platform (Fig. 2.12.16). This is a condition that can-
not be avoided. However, there should be some stated tolerance that
controls the mismatch dimension dr to minimize the turbulence
that is established.

E F G

a a

b b

dx
dx
dr dx
dr

F F
E G
Section 'b-b'
Section 'a-a'

Fig. 2.12.16TheFigure
effect of2.12.16
a root blending radius
on expansion
The effect of apassage discontinuity.
root blending radius on expansion
passage discontinuity.

Rotating portion circumferential arrangement


In the circumferential direction, efficient and reliable perform-
ance requires correct pitching of the vanes. Vanes cannot be proud
or recessed (see da values in Figure 2.12.17) from the mean inlet
and discharge edge beyond those tolerances established by the

211
Turbine Steam Path Maintenance and RepairVolume One

designer. These requirements will be considered in more detail later.


At this juncture it is sufficient to consider the two blade rowsone
stationary, and one rotating, as shown in Figure 2.12.17and con-
sider the following characteristics:

Vane setting angle . The vanes must be set in the cascade so


an expansion passage is formed between them. If the vanes are set
with an error in the stationary row s or rotating row r, the
shape of the expansion passage is compromised.

Vane pitch P. The tangential distance between adjacent pro-


files (stationary and rotating) must be set at the design value P
within the tolerances +dP and -dP. Variation of these pitches will
modify the shape of the expansion passage beyond design limits.

Vane axial placement a. The vanes must be set so the axial dis-
placement does not exceed the design values +da and -da. The
actual error at inlet and discharge may be different if the setting
angle is also in error.

Axial gap Ca. This gap between the rows can be compromised
if the values of +da and -da for the stationary and rotating rows
are beyond design limits.

Depending upon the method of manufacture, the vane settings


may not compromise the clearance Ca if there are lips at the inner
or outer diameters of the vane. This should be explored when these
characteristics are examined on the unit.

Diaphragm and stationary blade


manufacture and assembly
The axial positioning of the diaphragm is of critical importance,
as this establishes the relative positions of the stationary and rotating
portions of the unit relative to each other. It also establishes the axial
clearances between these components under all conditions. Figure

212
Steam Path Component Alignment and Stage Spatial Requirements

Fig. 2.12.17Section through a stage stationary and rotating blading, showing


the extent of spatial deformation and the possible effect on axial clearances.

2.12.18 shows the cross section through a diaphragm, in which the


axial position of all vertical (axial) faces is set by their position rela-
tive to face O-O. This face (O-O) becomes the steam seal sur-
face, and is located in the casing or blade carrier at a distance (A)
from some reference position, normally located within the cylinder.
All axial dimensions such as a, b, c, and d of Figure 2.12.19 are then
set by reference to face O-O, and are in a known position within
the total steam path, and relative to the anchor point of the casing.

Another important dimension shown in Figure 2.12.19 is the set-


back of the vane discharge edge from the b surface. This was dis-
cussed previously and is shown in Figure 2.12.17. At that time, the
possible effect in terms of the vane axial displacement from the
design condition by movements +da and -da on the axial clear-
ance Ca was considered. In the case of this diaphragm, it is clear

213
Turbine Steam Path Maintenance and RepairVolume One

that dimension b will not impact on axial clearance, but could


have an effect with some stage designs.

Also shown in Figure 2.12.19 are two radial seals that, on assem-
bly, will form a constriction above the rotating blade row. These are
machined to diameter Ds but must also be located axially to
achieve the correct axial position as shown in Figures 2.12.3 and
2.12.4.

O
a
Cro
O

Ca

Cri k

d
Figuredimensions
Fig. 2.12.18Control 2.12.18 for a
Control dimensions
diaphragm, for a diaphragm,
axial dimensions axial
shown relative
dimensions
to the steamshown relative
seal face to the steam seal
O-O.
face O-O.

The inner portion of the diaphragm will normally carry seal strips
to minimize leakage past the stationary blade row. These are shown
in Figure 2.12.1(a) and (b) as having a radial clearance Cri. At these
inner positions the rotor is in close proximity to the diaphragm, and
it becomes necessary to provide a chamfer at this location so that
clearance from the diaphragm to the rotor fillet radius and from the
wheel to central spindle portion is not reduced by operation. Such
chambers are shown as (d- k) on the discharge side.

214
Steam Path Component Alignment and Stage Spatial Requirements

O
A
a
O

Dot
b
Dit Ds

c
Dor

Dir k

W Dv
d

Fig. 2.12.19Portion of a2.12.19


Figure blade row showing the
seal strips
Portion of a at the diaphragm
blade row showinginner diameter
the seal stripsadja-
at the
cent to the rotor.
diaphragm inner diameter adjacent to the rotor.

Shown in Figure 2.12.18 is a detail of this region. The axial clear-


ance between the diaphragm inner web face and the wheel face is
shown as Ca. This clearance is shown without defining its radial
location. In the cold stationaryas a new conditionthis clearance
will be present along the entire radial face. It is also necessary to
ensure this minimum clearance is maintained around the fillet radius
of the wheel to the spindle section. It is known that certain operat-
ing characteristics will modify this value. It will change along the
radial direction being consumed to a greater extent at the inner
diameters and can also be affected by operating age.

Many rows of stationary blades are located directly into an inner


casing, as are many diaphragm rows. For reaction-type stationary
blades, this form of construction is used where the pressure drop on
the stages is such that the stages must be placed closer together to
conserve axial pitch, and where the pressure drop allows such a con-
struction.

215
Turbine Steam Path Maintenance and RepairVolume One

The inner casing or blade carrier of a unit locates within the


outer casing, and such a form of construction is shown in Figure
2.12.20. Again, contact is made at a seal surface O-O, and the
outer casing provides grooves into which this carrier locates. The
face O-O is at a distance A from some reference or datum posi-
tion, as discussed earlier.

Steam seal
face "OO".
O

f2 f3 f4

1 2 3 4
Dt4

t1 t2 t3 t4
b2 g2 g3 g4
b1 b3 b4 Dr4

Fig. 2.12.20A blade carrier locating Figure 2.12.20 blade rows, and showing the con-
four stationary
A blade
trol dimensions fromcarrier locating
the steam sealfour stationary
face O-O. blade rows, and showing the
control dimensions from the steam seal face O-O.

The rows of stationary blades carried by this casing have their


axial and radial positions defined by dimensions b1....b4 in the
axial direction, and g1....g4 and f1....f4 in the radial directions.
The blade rows also have an axial thickness defined as t1....t4. The
axial dimensions b and t define the axial space available to
admit the rotating blades and maintain an adequate axial clearance.
In the radial direction, dimensions g and f are selected to pro-
vide inner and outer diameters Dt and Dr for each stage suffi-
cient to achieve the design lap when the unit is assembled.

216
Steam Path Component Alignment and Stage Spatial Requirements

In many longer vortex design stages, the blade axial width


changes with radial position. In these designs, the axial clearance
may remain relatively constant, but be axially displaced (Fig.
2.12.21). This diagram represents a reaction stage with the stationary
blades located directly in a casing. Together with the width taper that
exists on the rotating blade and the coverbands, there are axial clear-
ances Cao at the outer diameter, and Coi at the inner. As shown
in the inset to this diagram, the width and overhang of the covers
bands can have an effect on the clearance and the m setting of the
rotors. The cover overhang (shown as da) is normally established
by final rotor machining.

Caulking

Casing

da

Cao Cao
Blade discharge

Lbo Det
height "h"

Dnt

Lbi
Cri
Cai
Der

Caulking

Fig. 2.12.21Showing the variation


Figureof 2.12.21
axial clearance Ca from the inner to
outerShowing
diameters
theinvariation
a vortexofdesigned stage. "Ca" from the inner to outer
axial clearance
di t i t d i d t

217
Turbine Steam Path Maintenance and RepairVolume One

Stage details
Modern units have many details incorporated into the individual
stage design that help improve or maintain stage performance, by
contributing to maintaining unit reliability or efficiency. Many of
these details are associated with sealing systems designed to reduce
wasteful leakage past the blade rows. There are too many such
details to discuss in great detail. However, their presence and func-
tion must be recognized. Not meeting the design requirements may
negate their value and degrade the total performance of the stage
and unit.

Seal production and configuration


Seals to limit leakage are arrangedwhere space permitsin
the steam path at those locations where a pressure drop exists and
the potential for losses occurs. Seals may also be incorporated in the
design where relatively small pressure differentials exist, but where
excessive steam flow past the area can occur to the extent disrup-
tions would occur in the flow pattern.

Consider the two stages shown in Figure 2.12.1(a-impulse) and


Figure 2.12.1(b-reaction), and the seals used in those stages. While
the overall stages may appear to be different, the following seals are
common to both designs, and perform the same function:

Radial seals of design clearance Cri below the stationary


blade row, forming a seal with the rotor body

Radial seals of design clearance Cro above the blade tip,


sealing on the coverband

Other radial seals are also shown. These include a root platform
seal strip (Fig. 2.12.5) with radial clearance Cr and an axial seal
shown on the impulse stageformed from the coverband of the

218
Steam Path Component Alignment and Stage Spatial Requirements

rotating blade row, and forming a constriction to the diaphragm


outer ring, clearance Ca.

In addition to these sealswhich are an integral part of each


stage designseals are used extensively at shaft-end positions in all
sections of a unit. This shaft-end packing is required to limit leakage
either of high-pressure steam from the steam path (contained within
the casing to atmosphere) or from the inlet between a high-pressure
section and a lower pressure section of a unit employing the reverse
flow design. In these reverse flow designs, the steam path of two dis-
tinct pressure range sections are mounted on a single rotor. In these
designs it is necessary to limit the quantity of steam leaking from the
high to the lower pressure section, which bypasses several stages
and does no work.

Seals are also required at the shaft-end positions of low-pressure


sections. They limit the quantity of air leaking into the unit, and min-
imize the quantity of sealing steam that must be supplied to each end
of the shaft preventing the inward leaking air from gaining access to
the low-pressure hood that is below atmospheric pressure.

Seal configuration can take various forms. The actual form


employed at any location is a function of the space and material
available for producing the seal, the most suitable method of attach-
ing it to either the stationary or rotating portions of the unit, and the
relative motion between them at that location.

The shaft-end seals are located at some known axial distance n


from the thrust or coupling face. For straight-through seals where
running interference cannot occur, this is the only dimension of con-
cern. However, if the seals are of the hi-lo type (shown for the rotor
portion in Fig. 2.12.8), the castellations are pitched at p, having a
set back distance w1 to the first castellation. These dimensions are
critical, as misplacement will often cause interference between the
castellations and seal strips.

219
Turbine Steam Path Maintenance and RepairVolume One

Seal geometry and arrangement


The geometry of typical seal systems should be considered for
installation at those locations where there is sufficient room for the
use of multiple series constrictionsfor a small number, two or
threeand where only single strips can be used. The possible
arrangements should be considered in both the axial and radial
direction, or in combination. At some axial locations (where the dif-
ferential expansion is large), it is possible the total number of strips
provided may not be effective. The particular strips that will be effec-
tive are dependent upon the differential expansion under a given set
of operating conditions. The larger number of sealing strips is pro-
vided to ensure that sealing at all axial positions and under all loads
is as effective as possible. The following possible arrangements for
multiple seal strips should be considered:

Straight through. The straight-through design (Fig. 2.12.22) has


minimum axial alignment requirements, though the only concern for
clearance is in the radial direction. The axial placement require-
ments are that the first tooth (k) is placed effectively at all extremes
of axial expansion conditions of the rotor, relative to the stationary
portions of the unit (and recognizing the requirement of the distance
n as defined in Fig. 2.12.2). It is also necessary that at rotor diam-
eter changes the space requirement for the last tooth (q) before the
diameter change is at a distance r from the shoulder, so interfer-
ence with the shoulder will not occur.

This form of labyrinth seal is used where axial movement can be


large from cold setting to final operating condition, and where large
axial movements are experienced due to differential expansion dur-
ing thermal transients. In this straight-through arrangement it is
known that a portion of the kinetic energy generated at each con-
striction is carried through from one seal strip to the next (Fig.
2.12.23). This carry-through effect is unfortunate, but unavoidable.

220
Steam Path Component Alignment and Stage Spatial Requirements

n
Cl q
k

Ds
Tooth
pitch

Figure 2.12.22
Fig. 2.12.22The straight through design with inserted teeth.
The straight through design with inserted teeth.

Fig. 2.12.23Showing the flow patterns set up in the


labyrinth seals.

Simple hi-lo configuration. It is normal to use the hi-lo type


configuration at those locations where differential expansion is not
large, and there are relatively small amounts of axial movement
under all extremes of transients. Such an arrangement is shown in
Figure 2.12.24, where alternate teeth form on the rotor and castella-
tions are produced on its surface. The principal advantage of this seal
is that thermal energy expended in expanding past a constriction
and converted to kinetic energyis almost completely destroyed
upon impact with either the castellation wall or the seal strip face.

221
Turbine Steam Path Maintenance and RepairVolume One

Cri Cro
Tooth
Pitch

Rotor
Ridge
Cold Width
Setting Ridge Pitch
Position "P"

Fig. 2.12.24The hi-lo teeth with Figure


a single 2.12.24
hi tooth at each castellation.
The hi-lo teeth with a single h tooth at each castellation.

The cold setting requirements for this design are shown in Figure
2.12.24, and the required axial clearances in the gap between
castellation shown in Figure 2.12.25. The axial clearances g and
h are set to design specification eliminating the risk of rubbing
interference between the teeth and castellation vertical surfaces.

h
g

Fig. 2.12.25Cold tooth


Figure 2.12.25
settings.
Cold tooth settings.

The radial clearances above the rotor and castellation are shown
as Cri and Cro. It in normal to make these clearances the same,

222
Steam Path Component Alignment and Stage Spatial Requirements

and this single clearance will be referred to as Cr. These radial


clearances are set by design at a minimum value sufficient to avoid
rubbing during operation, to allow the phenomena described ear-
lier, and possibly some rotor vibration, particularly during transient
operations.

The hi-lo configuration for large differential expansion. At


higher levels of differential expansionbut not so large that the
straight-through design need be usedthe design shown in Figure
2.12.26 can be applied. This design employs two high teeth for every
low one. With this design, only one hi tooth is effective on the
castellation at any time; the lo tooth is always effective. However,
at the extremes of operating position, there is always one hi tooth
effective. The cold settings are shown in Figure 2.12.26, and the
pitching requirements between the castellation are shown in Figure
2.12.25. However, the actual pitch values will have increased
because of the larger total pitch required between the castellations.
Compare Figures 2.12.24 and 2.12.26.

Cr

Tooth Pitch
Cold Ridge
Setting Width Ridge Pitch
Position "P"

Fig. 2.12.26The hi-lo configurationFigure 2.12.26


with two hi teeth to allow for large differ-
ential
Theexpansions.
hi-lo configuration with two hi teeth to allow for large differential expansions.

223
Turbine Steam Path Maintenance and RepairVolume One

The optimum point to convert from the use of the double hi-lo
to the straight-through design is a matter for design evaluation. In
Figure 2.12.22, there are shown eight effective teeth, and in Figure
2.12.26 there are nine effective teeth (only one hi tooth is effective
at any time). It is a matter for the design evaluation to establish which
will be the most effective design to use in the axial space available,
recognizing the superior effectiveness of the hi-lo configuration in
destroying the kinetic energy of the leaking steam.

The seal teeth shown in Figures 2.12.22 through 2.12.26 are of


the inserted strip type. However, a commonly used form of tooth is
that which is carried on an inserted gland ring (Fig. 2.12.27). These
forms of sealing arrangements are considered in detail in chapter 11.

Stationary
portion

Rotor Cr
Dd

Fig. 2.12.27The segmented


Figure seal ring, with
2.12.27
control dimensions.
The segmented seal ring, with
control dimensions.

The selection of seal configuration


The seal configuration used at any location is dependent upon a
number of factors acting to make one arrangement more suitable
than others. Among these considerations:

224
Steam Path Component Alignment and Stage Spatial Requirements

The differential expansion at that location together with any


other factors that might influence the axial position of the
rotor relative to the stationary portion of the unit. This can
include certain requirements including the need to achieve
axial shift for coupling adjustment of component disassembly

The axial space that is available to mount the seals. There is


no advantage to using more seals than can reasonably be
effective. To place the seal strips too close will make them
ineffective

The cost of producing the seals. If the financial cost of pro-


ducing and mounting seals is more than the cost that can be
achieved, then the less effective sealing system will be
installed

Consider the possible geometries and arrangements used at any


location. There are three major areas to consider. These include:

At shaft-end seal positions. At shaft-end positions, the most com-


monly used seal at the high-pressure, high-temperature locations
close to the thrust block is that shown in Figure 2.12.24. That is the
simple Hi-Lo arrangement. At low-pressure locations where the
differential expansion is considerably greater, the straight-through
arrangement is most common. Each axial location is analyzed dur-
ing the design process, and the most effective design consistent with
manufacturing and fuel costs is chosen.

Under the diaphragms and stationary blade rows. In the high-


pressure, high-temperature stages of impulse units (the seals at the
diaphragm-to-rotor locations), the seal system is selected based on
the same considerations as those governing the selection of the sys-
tem used at the shaft-end positions. Under these higher condition
diaphragms, there is normally sufficient axial space that multiple con-
striction seals can be used, and because of the high-pressure drop
across them, it is necessary such systems be used to limit leakage.

225
Turbine Steam Path Maintenance and RepairVolume One

Similarly, in reaction stage stationary blade rows it becomes nec-


essary to provide seals at the inner diameter, between the inner
blade tip and the rotor surface. The inner diameter will normally
have an inner coverband, which is either attached by the use of a
tenon formed on the vane, or with a coverband formed integrally
with the vane. Figure 2.12.28 shows four possible systems for attach-
ing such seals at this inner location. In (a) and (b), the seals are
shown as part of an inner coverband that is itself attached to the
blade tips using tenons. In (a) the seals are formed integrally with the
cover and in (b) they are caulked in place. In (c) and (d), the inner
coverband is formed integrally with the blade vane and the seals are
mechanically attached. In (c), these are attached by caulking and in
(d) by an attachment method that normally is welding or brazing. In
each of these four cases, normally seals require trimming after
assembly of the blades to the casing or blade carrier to achieve the
seal diameter and clearance Cri.

Cri

Rotor (a) (b) (c) (d)


Surface

Fig. 2.12.28Showing alternate seal Figure 2.12.28 at the inner diameter of stationary
arrangements
blades.Showing alternate seal arrangements at the inner diameter of stationary blades.

At rotating blade tips. The possible arrangements and design


options at the rotating blade tips are far more varied than at the other
locations. The options available allow these seals to form axial or
radial constrictions, or a combination of both. Typical seal systems
at the rotating blade tips are shown in Figures 2.12.3 and 2.12.4.

226
Steam Path Component Alignment and Stage Spatial Requirements

The seal strips used can either be produced integrally with the
blade vane cover or be inserted into a portion of the unitrotating
or stationary. It is not common with rotating blades to caulk attach-
ing seals to the rotating portions because there is the possibility of
centrifugal effects detaching them during operation. Such designs do
exist and have been used successfully, but considerable care is
required in the caulking operation. It is most common for the seals
to be caulked to the stationary portions of the stage.

From a maintenance and economic perspective, it is preferable to


form the seals from separately attached components that are caulked
or inserted into the stage. This allows their replacement when they
have become worn (causing a larger leakage area and no longer seal-
ing). Separate attachment is considered optimum, because this allows
the seals to be changed without the need to replace the entire major
component of which they are a part. Although procedures are now
available to allow worn seals to be weld-rebuilt, the cost is higher
than simply replacing inserted seal strips.

v v v v

Cro Cro
Cro Cro
(a) (b) (c) (d)

Fig. 2.12.29Possible seal configuration at a2.12.29


Figure blade tip.
Possible seal configuration at a blade tip.

Figure 2.12.29(a) shows radial seals at a rotating blade tip. These


seal strips have been produced integrally with the blade cover,
which can be attached or formed integrally with the blade vane. This
type of seal is produced and the final form is machined after assem-
bly to the rotating blades. Figure 2.12.29(b) shows a design in which

227
Turbine Steam Path Maintenance and RepairVolume One

seal strips are caulked into the outer surface of the coverband. As
previously stated, this form of seal is not common, and can be a
source of damage if the steal strips detach. Figure 2.12.29(c) shows
a radial seal in which the seal strip is caulked into the stationary por-
tion above the coverband. Figure 2.12.29(d) shows a seal in which a
segment is inserted into a special groove produced in the casing. In
each design shown, the axial control distance from the reference
face is shown as v and the clearance as Cro.

Figure 2.12.30 shows an arrangement that provides a seal in the


axial direction. This is possibly the most common form of seal for
impulse stages, in which a knife-edge is produced integrally with the
coverband inlet edge. This type of seal is normally trimmed after
assembly of the coverband to the blade tips, and establishes the axial
clearance Ca. This form of seal may not necessarily be referenced
to a datum face, when the coverband is changed on maintenance.

Casing

Diaphragm
outer ring

Rotating
Stationary
blade blade

Ca

Fig. 2.12.30An axial2.12.30


Figure seal at the blade tip.
This seal is
An axial formed
seal at theintegral with
blade tip. theseal
This cover
is
band.formed integral with the cover band.

Figure 2.12.3 and 2.12.4 shows that there are designs combining
axial and radial seals. There are many acceptable arrangements, and
it is necessary to consider the spatial arrangements in terms of the
particular design.

228
Steam Path Component Alignment and Stage Spatial Requirements

BLADE VANE AND


CASCADE GEOMETRY
The method of calculating or determining the form blade vanes
take is complex. In a future chapter we will review vane form from a
theoretical perspective, making the assumption that the design spec-
ified form can be met without exceeding design specified tolerances.
There are, however, certain characteristics of the vaneits form and
placement on the root platformwhich have a significant effect on
the performance of the total unit. These geometrical requirements are
considered in this section, together with efforts made to define which
of these characteristics are important and cannot be relaxed without
causing a deterioration in the performance potential of the unit. This
section will also explain which parameters defining the total blade
vane need to be considered when blades are manufactured.

Vane form requirements


The vane form is not critical in and of itself. The actual form of a
blade is important only in terms of the shape of the expansion pas-
sages it produces between adjacent elements. In assembling blades
(either stationary or rotating) errors in the profile form (or in the rate
of change in vortex designs and its placement) can influence two
expansion passages, and has therefore the potential to degrade per-
formance.

Profile shape. The expansion passage must be considered in terms


of the demand for and degree of convergence required. In supersonic
rows, it must be recognized there is a need to produce a converging-
diverging passage form. However, no thought has been given to the
possible effect that manufacturing tolerances and errors could have on
the passage, or even where tolerances need be applied. The profile is
designed to achieve certain technical requirements. Among these are

229
Turbine Steam Path Maintenance and RepairVolume One

the needs to form an acceptable passage, and expand the steam with
a minimum of losses due to form. The shape of the profiles will also
control the total discharge area from the row. This will be considered.

In order to achieve its functions as effectively as possible, there


are certain characteristics of form that must be observed. There are
parameters that are defined by design, which specifies shape, and
assign a tolerance band within which they are acceptable.

Fig. 2.13.1A blade mounted in a shutter


gauge for profile confirmation.

230
Steam Path Component Alignment and Stage Spatial Requirements

It is normal for the design specification to define an envelope of


tolerances for the profile, providing a plus or minus (+/-) tolerance
band for various locations on the surface. What such a tolerance
band cannot establish is the extent to which such a tolerance can
vary within the envelope. It is possible for the band to be (+) at one
location and (-) at another adjacent location. Assessment of this con-
dition calls for mature judgment on the part of the design engineer.
Gauging the acceptability of a profile is normally undertaken using
a shutter or guillotine gauge (Fig. 2.13.1) and a typical envelope
of tolerances (Fig. 2.13.2) where variations from the defined shape
are shown as dT and -dT. Special tolerances are applied to the
discharge tail.

+dT

+dT -dT

-dT Rmax

+db
b

+db
b

Fig. 2.13.2The envelope of tolerances defining the profile form.


2.13.2

231
Turbine Steam Path Maintenance and RepairVolume One

Chord C and thickness T. From mechanical considerations,


the more critical considerations of the profile are the chord and the
vane thickness. These are best defined as the maximum thickness
along a mean line traversing the median thickness (Fig. 2.13.2 and
2.13.3), where the maximum thickness is 2.Rmax, measured near
the center of the profile.

Rmax

Ct

Rmax
Cf

Ch

Figure
Fig. 2.13.3Definitions of 2.13.3
profile thickness and
chord.Definitions of profile thickness and chord.

The chord of the vane can be given one of several definitions but
the most appropriate depends upon the matters being considered. If
discussion of any experimental work is involved, the definition used
to derive any particular set of results is the most appropriate. The fol-
lowing definitions can be applied:

as the distance Ch across the horns of the profile. This eas-


ily measured characteristic can be determined with relative-
ly unsophisticated measuring instruments

as the distance along the center of the profile Cf. To deter-


mine this value, it is necessary to have it defined from com-

232
Steam Path Component Alignment and Stage Spatial Requirements

puter calculation. If only the profile is available, constructing


the position of the center from drawn radii and then measur-
ing must establish the chord

as a line drawn from the center of circles defining the inlet


nose and discharge tail Ct. This is not a common definition
but is most appropriately used when methods of calculating
profiles determine these are positions to aid in the drawn
construction of the profile

These parameters C and T influence the mechanical strength of


the profile, since they influence both the profile area and the section
modulusa critical consideration when evaluating vibration charac-
teristics.

Inlet nose. A profiles inlet nose has the potential to modify the
streamline form of the steam as it passes between the vanes of a
blade. Consider the three profiles shown in Figure 2.13.4. The cen-
ter element (S) has an inlet nose that divides the incoming steam
and diverts it to the left or right to flow through one of two passages
(R-S or S-T) that it helps form. Steam entering the row with a high
relative velocity (W1) introduces a stagnation region (s).

If the vane is manufactured to design there will be no disconti-


nuities in the nose form, and the flow will be even in the passages
(other than the possible separation bubble that can form). However,
vane manufacture is a process that can involve a certain amount of
hand polishing, and any uneven surfaces produced (as shown by flat
y-y in Fig. 2.13.5) can magnify the disturbance that occurs and
introduce significant turbulence into the flow.

The vanes shown in Figure 2.13.4 are typical of the rotating


blades in an impulse type design. For stationary blades the velocity
of the entering steam C2 (see Fig. 2.4.15) is much lower, and will
therefore cause less degradation in the stage efficiency.

233
Turbine Steam Path Maintenance and RepairVolume One

Fig. 2.13.4Flow division at inlet to the stage.

Discharge tail. The discharge tail helps form the discharge point
from the expansion passage. As such, it must provide the correct dis-
charge area and angle (as explained previously). Stressesboth
bending and direct tensileare high in the tail region, and it is nec-
essary to ensure the definition of tail thickness is observed. It is nor-
mal that the envelope of tolerances at the tail is limited to design
with a (+) tolerance only. No undersize thickness is acceptable.

The tail can be finished to several forms as shown in Figure


2.13.6. In (a), the flat is maintained to a knife-edge. In fact, there is
always a degree of rounding [as shown in (b)] and at times a com-
plete rounding, as shown in (c). To assist in measuring discharge tail
thickness, the most convenient method is to measure at some speci-
fied distance (d) back from the discharge point to establish the
thickness (b), as shown in Figure 2.13.7. Both d and b can be
defined by design sufficient to permit accurate determination, allow-
ing for the fact that the distance d is dependent to a considerable
extent on the polished form of the discharge tail. Measuring too

234
Steam Path Component Alignment and Stage Spatial Requirements

close to the discharge point can result in measurements made on the


radii formed by polishing, i.e., q-q.

Fig. 2.13.5A flat produced on


the inlet nose of a vane, modify-
ing the form of the profile.

Fig. 2.13.6Details of the discharge tail,


and various forms.

235
Turbine Steam Path Maintenance and RepairVolume One

Fig. 2.13.7Quantifying discharge


tail thickness.

Vane setting angle . If a vane is manufactured correctly


meeting all requirements of form and finishit is still unable to form
an acceptable expansion passage if it is set within the cascade in
such a manner that it is incorrectly inclined in the axial/tangential
direction. As shown in Figure 2.13.8, to be correct, this setting angle
must have its inclination . Modifying this angle by unacceptable
amounts will modify the shape of the passage, anddepending
upon the point in the profile about which the vane rotatessuch
work can modify the rate of convergence and modify the discharge
throat by unacceptable amounts.

As can also be seen in Figure 2.13.8, errors in the setting angle


can influence the effective width of the vane, causing errors of
+dW or -dW. The increase or decrease in width can occur at
either of the edges, depending upon the point in the profile about
which the untwist has occurred.

In service, this setting angle on vortex blades will become mod-


ified by the untwist of the blade vane occurring as a consequence
of the centrifugal stiffening effect. The extent to which the blade will
untwist is influenced by the stage hardware, such as the coverbands
and tie wires that help restrain this effect to a certain extent.

236
Steam Path Component Alignment and Stage Spatial Requirements

Fig. 2.13.8The effect of setting angle errors d on row


width dW.

Pressure and suction faces. In the portion of row cascade shown


in Figure 2.13.9, three profilesR, S and Tenable steam to
enter the row at a velocity W1 and discharge with a velocity W2.
Each vane forms one surface of two adjacent expansion passages
(the suction face of one and the pressure face of another).

Figure 2.13.9 shows three profiles, or faces. The center profile


(S) indicates these faces. The concave surface of the profile (S)
has a positive pressure relative to the mean pressure (Pmean) in the
passage developed on it due to the change in steam momentum.
Similarly, the opposite face in the passage has a pressure negative to
the mean developed on it. Therefore, the total force developed on
the profile is the sum of the forces on both sides of the profile. This
is shown in the pressure diagram, where pressure through the row
reduces from Pi at inlet to Po at discharge.

237
Turbine Steam Path Maintenance and RepairVolume One

+dPi
W1
Pi

Suction Suction
face face

Pressure
face

Pmean
R S T

W2 Po -dPo

Fig. 2.13.9The pressure and suction


Figurefaces of a profile, and representation of
2.13.9
the pressure distribution
The pressure andinsuction
the expansion
faces ofpassage.
a profile, and representation of the
pressure distribution in the expansion passage.

Vortex form and profile change. The vane profile is a two-


dimensional form. However, the vane of a blade is a three-dimen-
sional body, whichdepending upon the design philosophy used to
establish vane requirements for that rowcan have a profile that
varies as a function of its radial position. This is done to accommo-
date the changing steam velocities and relative inlet angles to the
row. Figure 2.13.10 shows the stacking diagram of a vortex blade,
including the profile change at equidistant radial positions above the
root section. The relative position shown for the profiles above the
stacking point is selected to minimize the centrifugal bending
stresses during rotation.

Various vane angles are essentially defined or selected by the


thermal design process, and so in selection of profiles, the designer
must meet and ensure the expansion efficiency guaranteed in the
quotation to the client. Stage profiles are shown in Figure 2.13.11.
The important angle considerations follow.

238
Steam Path Component Alignment and Stage Spatial Requirements

Fig. 2.13.10The Stacking Diagram for a vortex vane.

Vane inlet angle 0 or 1. The vane inlet angle is selected


to allow the steam from the previous row to enter the expansion pas-
sages with a minimum of incidence. The inlet nose of the majority of
modern profiles is rounded. Experimental work has shown that larg-
er inlet radii (Rs and Rr of Fig. 2.13.11) are better able to accom-
modate incidence with an off- design inlet angle. However, there are
losses associated with all such incidences.

For the stationary row of an impulse design, the velocity entering


the row is considerably less than that entering the rotating row.
Thereforebecause losses are velocity sensitivethese stationary
rows are able to accept greater incidence than the rotating rows. In
high-reaction stages, velocities entering the stationary and rotating
rows are of the same general magnitude. This means that while both

239
Turbine Steam Path Maintenance and RepairVolume One

suffer the same order of losses due to incidence, the entering veloc-
ity in a reaction row is somewhat smaller than the impulse stage, and
the level of losses are smaller.

Certain older designs of vaneparticularly those used in


impulse stage rotating elementshave an inlet that is not rounded,
but flat (Fig. 2.13.6). In these stages, the flat surface will interfere
with the inflowing steam, deflecting a portion and causing a degree
of turbulence that is carried into the expansion passage. As a unit
ages, it is possible this flat will become partially rounded, but there
will still be some level of loss.

Rs 0

s
Rr

1 1
r

Figure
Fig. 2.13.11Stationary and2.13.11
rotating blades showing
basic characteristics.
Stationary and rotating blades showing
basic characteristics.

Vane discharge angle, 1 or 2. The vane discharge angle is


shown as the mean angle of inclination of the vane discharge tail.
The value used in determining the angle requirement at discharge is

240
Steam Path Component Alignment and Stage Spatial Requirements

the inverse sine of the ratio O/P. However, the discharge portion of
the vane must be set to support these geometrical requirements, and
while there are minor differences, (as will be discussed in a further
section concerning the pitch) the vane tail inclination assists in con-
trolling the throat for any vane pitch at discharge.

Vane turning angle . The vanes must be designed to admit the


steam at an angle 0 and discharge it at an angle 1 on a sta-
tionary element, admit steam at an angle 1 and then discharge it
at an angle 2 on a rotating element. This means the steam must
be deflected through an angle (known as the turning angle). This
angle is given the designation , being s for the stationary ele-
ments and r for the rotating.

The magnitude of these turning angles can be established from


the relationships:
For the stationary row: s = 180 - (o + 1)
For the rotating row: r = 180 - (1 + 2)

In many stages, the profile actually used may not meet precisely
the requirements of inlet, discharge, and turning angle s or r.
This is because profiles are expensive to design, and situations often
require that existing profiles be employed rather than newly devel-
oped for any specific application.

Axial placement
Figure 2.12.17 shows the theoretical layout of the vanes for both
stationary and rotating rows of a single stage. This figure also indi-
cates that vanes can be proud by an amount +da or recessed by an
amount -da. Before discussing these errors, it is necessary to estab-
lish a definition for plus (+) and minus (-) deviation from the theoret-
ical line of the vanes. A suitable nomenclature that will be used is:

241
Turbine Steam Path Maintenance and RepairVolume One

in a positive deviation, the vane projects beyond the con-


straints of the theoretical boundary line. This can represent a
movement up or down stream in the axial direction,
depending upon whether the row being considered is sta-
tionary or rotating, and whether this is at the inlet to or
discharge from the expansion passage

in a negative deviation, the profile is recessed from the theo-


retical boundary line. This negative movement, like the pos-
itive, is a function of the vane and its position relative to
steam flow direction

The inlet and discharge edges are defined as the theoretical posi-
tion of the vanes at inlet to and discharge from the row. These are
shown for a stationary row in Figure 2.13.12, where both proud and
recessed edges are shown at the design limits of +da and -da.
This figure indicates that if the blades are set with the correct setting
angle, a blade proud to one edge is recessed at the other. However,
errors in the setting angle may cause both edges to be either proud
or recessed.

Fig. 2.13.12Blade cascade showing proud and recessed blades at the inlet
and discharge edges.

242
Steam Path Component Alignment and Stage Spatial Requirements

Inlet and discharge edge. The inlet and stationary edges shown
in Figure 2.13.12 define the design-specified axial limits for the row.
Also shown are the positive (+) and (-) limits at inlet to and discharge
from the row. The limit for a proud or a recessed edge is set so the
stage will perform at the design-specified level. The tolerances are
set dependent upon the methods of manufacture.

Stationary blade elements. The limits established for stationary


blades are normally set at values a little more liberal because of the
manufacturing methods. This is particularly necessary when vanes
are located in cast or fabricated rings. These blade elements can also
be adjusted to achieve an improved geometric arrangement.

Rotating blade elements. The rotating blades are normally set


with limits that are more restrictive than those applied to the sta-
tionary row. This is necessary because misplacement of the profiles
will induce stresses in the vane due to bending as the blades rotate.

In general, the stationary rowparticularly for fabricated dia-


phragm stageshas a greater tendency for error because of their
method of manufacture.

Tangential placement (pitch)


The design process sets the tangential placement of one profile
relative to others in the row, so the throat is produced on the suction
face of the adjacent element. For curved-back profiles, tolerances
may need to be more restrictive because of the changes errors intro-
duce in the discharge angle.

The cascade
The cascade is a series arrangement of blades that are placed
forming a series of expansion passages. This arrangement is selected
to achieve two concurrent objectives:

243
Turbine Steam Path Maintenance and RepairVolume One

To provide a discharge throat at the tail position. This throat,


when integrated along the radial height of the vane, provides
a discharge area for the single throat that when integrated
around the row provides the total discharge area required for
that row

To provide the discharge angle required to direct the expand-


ing steam into the following blade row

To produce a satisfactory cascade fulfilling the requirements of


area and angle the following geometric and spatial characteristics
must be met:

Vane placement. The vane must be placed and secured so its


position within the cascade is correct and meets design require-
ments. Shown in Figure 2.13.13 is a rotating blade pair, of a typical
high reaction stage, and the development of the passage width as a
function of the axial position. This represents the design optimum
form of the expansion passage, and should be met within close tol-
erances.

P O

Fig. 2.13.13Rotating blade Figure


pair for a2.13.13
high reaction stage showing the
developed
Rotatingexpansion
blade pairpassage.
for a high reaction stage showing the developed
expansion passage.

244
Steam Path Component Alignment and Stage Spatial Requirements

Vane twist. The thermodynamic optimums of a blade row are a


function of blade height, therefore the vane should change its geom-
etry along the radial height to accommodate these variations and
minimize losses.

In many small radial height stages, there is often insufficient pre-


dicted variation of the inlet, discharge, and turning angles for the
cost of twisted or vortex blade to be justified. The vortex blade is, in
general, more expensive to manufacture, but fuel cost saving can
often easily justify the additional expense of producing the vortex
profile. (However, with multi-axis milling, this production cost dif-
ferential is relatively small, and only the complexity of milling small
radial-height blades with integral covers will induce suppliers to use
the cylindrical form.)

A general, but not absolute rule is if the ratio of vane height to


mean diameter (h/Dm) is greater than 0.356, then a vortex vane
should be used. However, many manufacturers can justify the use of
a vortex vane from a cost/performance perspective when the h/Dm
ratio exceeds 0.10 to 0.15, depending upon fuel costs and predicted
unit load factor.

Expansion passage form. Figure 2.13.13 shows the passage form


between adjacent vanes. In this cascade, the throat width reduces
from inlet to a throat of O at discharge. The pitch at discharge is
P.

The ideal passage shape is designed to turn the steam through


the correct angle, accepting the inflowing steam at an angle 1
and discharging it at an angle 2. The passage should diverge at a
rate sufficient to form the correct discharge area. It is also necessary
to ensure there are no sudden changes in the radius of curvature of
either the pressure or more importantly the suction surface, as this
effect can cause premature separation of the boundary layer.

245
Turbine Steam Path Maintenance and RepairVolume One

Throat. The throat between the discharge tail and the suction
surface of the adjacent blade is one of the more critical characteris-
tics defined by the design process. The product of the mean throat
along the radial height of the vane and its height will give the dis-
charge area from the passage. The sum of all the individual throat
areas will give the discharge area of the stage that in turn defines
velocities and pressures at discharge from the stage.

Pitch. The pitch is the circular distance between adjacent pro-


files, and if there are Zb profiles in a row, the theoretical pitch
shown as P in Figure 2.13.13 can be found from:

.D
P =
Zb

where:

D is the diameter at which the pitch is being considered

Note: When calculating pitch, it is normal to use the circular


pitch as determined from this equation. However, even when a rel-
atively small number of blade elements are employed, the chordal
pitch is often used, as it is more accurate.

Throat/pitch. The ratio of throat/pitch (O/P) establishes the


steam discharge angle from the passages. In many rows, an audit is
made of the values of O and P after assembly (plus any needed
adjustment), and tolerances are set on these values.

Typical of the tolerances applied to rows are:

Parameter Tolerance Range


Throat O 4.0 - 5.0 %
Pitch P 4.0 - 5.0 %
Throat/ Pitch O/P 2.0 - 2.5 %

246
Steam Path Component Alignment and Stage Spatial Requirements

Actual tolerances for any stage are dependent upon whether they
are stationary or rotating, vane radial height, and method of manu-
facture.

Discharge area Ad. The total discharge area from any row
establishes the pressure at the discharge point and, therefore, the
energy on the stage. In establishing this total area, consider the area
of the individual passages (ad). Figure 2.13.14 shows the variation
of throat (O) along the radial height of a vortex blade vane (either
stationary or rotating). It can be seen from this figure that the throat
does not vary linearly, but is some function of the vane twist and set-
ting angle variation. Therefore, the discharge area of the individual
throat is:
h
ad = o o.dh

and the total discharge area Ad for the row is:


z
Ad = o ad

Ot
Height

Om

Throat
Or

Fig. 2.13.14Showing the radial vari-


Figure 2.13.14
ation of throat
Showing on a vortex
the radial blade.
variation of throat
t bl d

247
Turbine Steam Path Maintenance and RepairVolume One

Swallowing capacity. A blade row consists of a number of ele-


ments, each placed relative to adjacent profiles sufficient to complete
the 360 or design-specified portion of the total arc for some station-
ary rows. If each vane is correctly placed, then each of the expansion
passages will accept an equal amount of the fluid in the row
expressed as the total flow divided by the number of expansion pas-
sages comprising the row. The quantity of steam admitted to individ-
ual passages is defined as the swallowing capacity of that passage.
The sum of the swallowing capacity of the individual passages com-
prising a row is the total flow in the row less any quantity of steam
that bypasses due to leakage. This is taken up in the following section.

THE EFFECT OF VANE


PLACEMENT ERRORS
We have considered the blade cascade and the characteristics
required to define it, as well as how the parameters of the cascade
are able to influence the total performance of the stage and directly
impact the efficiency of energy conversion.

However, because various manufacturing techniques are


employed to produce steam path componentsand because of
varying engineering tolerances available by design to allow the
components to be produced at acceptable costsvariations in axial
setting (a), pitch (P), and setting angle () are expected as
part of a normal manufacturing process. Variations within design tol-
erances are anticipated, and the design should have sufficient mar-
gin in the predicted performance so the unit can reach guaranteed
efficiency levels.

It is inevitable that components will be manufactured and assem-


bled exceeding these tolerances. It is the responsibility of the design

248
Steam Path Component Alignment and Stage Spatial Requirements

function to evaluate and recommend corrective actions, if these are


considered necessary. It is also possible operating problems will
modify the geometry of the stages and stage components. When it is
available for inspection it is normally the responsibility of the oper-
ating engineer to consider these and their potential effect on the unit.

Various forms of errors can be introduced into the steam path,


and it is necessary to consider the effect and extent to which these
are acceptable. Design tolerances of the profile and the cascade
must be based upon what can be tolerated before the quality of the
cascade energy conversion ratio deteriorates by an unacceptable
amount. Tolerances must not be set to accommodate manufacturing
techniques and what can be achieved. If a particular technique can-
not produce adequate accuracy, the technique must be changed
not the tolerances.

Unfortunately, there are no simple rules that can be formulated


to determine acceptability. The extent to which errors are acceptable
depends upon a number of factors, including the form of the profile,
the turning angle, and a number of intangible factors specific for the
row. The three most important are discussed below:

Axial placement error


The effect of axial placement has been considered, and Figure
2.13.12 shows proud and recessed blade vanes within both the sta-
tionary and rotating blade cascades. For both rows, it shows accept-
able tolerances of +da and -da. What have not been considered
until now are the consequences of these errors on the parameters of
the cascade, and what would cause such errors to occur. The most
common conditions to induce errors follow:

Vane position on the root platform. A rotating and some sta-


tionary blades can be considered to consist of two principle por-
tionsthe vane and the root. They are connected through an inte-

249
Turbine Steam Path Maintenance and RepairVolume One

grally connected portion known as the root platform. It is necessary


during the manufacturing phase to ensure the vane is located on the
root platform in the correct spatial position relative to the adjacent
elements, forming an acceptable expansion passage.

Figure 2.14.1 shows a vane drawn relative to the root platform.


If it is assumed the platform is located correctly above the root load
bearing or load transfer surfaces, then the vane must be positioned
so the distances dy1 and dy2 from the vane inlet and discharge
edges are correct. Therefore the vane is axially correct. In these posi-
tions the vane will have a total width of Wv, and the root platform
a total width of Wr. It is possible that on many roots the distance
Wv is equal to Wr. Errors can still be encountered, though nor-
mally to a lesser degree.

dy1

Wr
Wv

dy2

Fig. 2.14.1Vane profile


Figure placed on a
2.14.1
lozenge type root
Vane profile platform.
placed on a lozenge type
root platform.

If however, the vane S is axially misplaced by an amount e


(shown in the partial cascade of Fig. 2.14.2, with three vanes: R,

250
Steam Path Component Alignment and Stage Spatial Requirements

S and T), and the neighboring vanes R and T are correct,


then the expansion passages formed will be incorrect. Throats
formed with both adjacent vanes Ors and Ost are incorrect.
Their total discharge throat (and therefore, their area) may be the
same and within design tolerances, but their discharge angles will be
modified (because the throat has changed) while the pitch remains
substantially correct.

R S T

Wv
Os
Or
s

Figure
Fig. 2.14.2Three vanes R, 2.14.2
S and T placed in cascade, the
Three
center vane vanes
S beingR, S and T
misplaced placed in cascade, the
axially.
center vane S being misplaced axially.

For stationary elements, it is possible a larger tolerance can be


accepted in terms of the vane movement from the theoretical axial
position (inlet and discharge edges), provided all elements in the row
are affected or in error to the same degree. However, with rotating
elements, excessive axial tolerances cannot be applied and accept-
ed, even if all elements have the same degree of error and would
produce an acceptable expansion passage with adjacent elements
on assembly. These more stringent tolerances on rotating elements
are necessary because vane displacement on the root platform will
introduce a centrifugal bending moment into the blade, leading to

251
Turbine Steam Path Maintenance and RepairVolume One

increased root and vane stresses. This effect is shown in Figure


2.14.3, where the center of gravity of the vane Gv (a distance Y
from the discharge edge) has been axially displaced on the root plat-
form by an amount dY. This introduces a bending moment M.dY,
where M is the total centrifugal load of the vane.

dy

Gv

Fig. 2.14.3Showing vane center of


Figure 2.14.3
gravity
Showing vane centerfrom
Gv displaced the root
of gravity Gv
platform.
displaced from the root platform.

Root form position. The axial relationship of the root load-bear-


ing surfaces to the center of gravity of the vane G is shown for a
tangential root form in Figure 2.14.4. A theoretical line (G-Gv) is
shown passing through the center of the root profile and then
through the vane Gv -G. In Figure 2.13.10, it is shown that the
vane profiles are stacked above their common center of gravity
(G). This position (G) lies on the radial line G-G, shown in
Figure 2.14.4 as Gv.

In the manufacturing of the blade, the root form is produced with


distances from the line G-G to the three pairs of load-bearing sur-

252
Steam Path Component Alignment and Stage Spatial Requirements

face center points. Lul to Llr is the position at the upper u, mid-
dle m, and lower l radial positions on the left l and r of the
centerline. These distances must be controlled, which is relatively
uncomplicated as the root form is cut with a form cutter producing
all six surfaces simultaneously. What must be assured, however, is
that the vane is positioned correctly, at a maximum distance speci-
fied as dax.

Wvt

dy1 dy2
Wr
dax
Gv
Wvr

Lul Lur

Lml Lmr

Lll Llr

Fig. 2.14.4Showing
Figure 2.14.4 the relation-
ship between
Showing the the vane andbetween
relationship root load
bearing surfaces
the vane to aload
and root line bearing
through
the centertoofagravity.
surfaces line through the center
of gravity.

This correct root profile placement is assured by ensuring the


root form cutter is positioned correctly as it passes through the root
block. This root form is normally produced before the vane is

253
Turbine Steam Path Maintenance and RepairVolume One

machined and in fact is used to locate and hold the material block
during vane machining. However, if machining the vane first pro-
duces the vane portionwhich can happen with envelope or preci-
sion forgingsthen care must be exercised in positioning the vane
before cutting the root.

Wvt
G
Daxt

dy1 dy2
Wr
daxr
Gv

Wvr

Lul Lur

Fig. 2.14.5A blade2.14.5


Figure with the vane
tilted in the
A blade axial
with the direction
vane tiltedbyinan
the
amount Dax by
axial direction at the tip section.
an amount "Dax" at
the tip section.

Vane axial lean. In addition to vane misplacement on the root


platform, it is possible for the vane to be inclined in the axial direc-
tion. This effect is shown in Figure 2.14.5, where the vanehaving
the same axial location at the root section and being offset in the

254
Steam Path Component Alignment and Stage Spatial Requirements

axial direction by an amount daxis also tilted at an angle ,


giving a total offset at the tip of Daxt. The root offset is defined as
daxr and the tip movement in the axial direction is daxt-daxr. The
net effect of this tilt can be to make the root axial position entirely
satisfactory at the root section but beyond limitsor certainly with a
greater total axial displacement at the tip.

This axial lean will also produce a bending moment in the blade
increasing stress levels. Depending upon the extent, axial lean will
induce higher stress levels in the blades and, in the majority of
vanes, will tend to cause a greater level of distortion to the expan-
sion passage forms. Each leaning blade affects two passages.

Vane untwist. As the unit starts up, the speed and centrifugal
loading of the vane increases. Figure 2.14.6 shows two sections of a
vane mounted in their final manufactured position above the center
of gravity (or stacking position G at axial A and tangential T
locations). The effect of rotor rotation is to make the vane untwist
in a counterclockwise direction (M), so the tip section Ti rotates
about G in an effort to achieve a position above the root section
Ro. If the vane is unconstrained, the extreme position of the tip
section discharge point would move from coordinates X, Y by
amounts dX, dY. In untwisting, the vane will experience a turn-
ing moment in the vane equal to M. This untwisting will modify the
tip form, causing a modification of the throat and therefore the
throat/pitch ratio that defines the discharge angle.

With freestanding blades, the only constraint to this modification


of section position is the resistance of the blade material to the bend-
ing moment set up by vane rotation. The untwist is calculable. The
vane is therefore calculated with pre-twist built in, so that during
operation the profiles along the vane height will rotate to the posi-
tion required by design to give the throat distribution required as a
function of height.

255
Turbine Steam Path Maintenance and RepairVolume One

Fig. 2.14.6A vane with tip intwist, about its


center of gravity.

For those blades with coverbands (or tie wires) at their tip
attached to the vanes in such a manner they resist untwistthe situ-
ation is somewhat different. The untwist will experience consider-
able resistance and the vane will move only by relatively small
amounts. Therefore, the blade can be manufactured with the throat
distribution required during operation.

Coverband distortion. As we just mentioned, the coverband can


resist untwist of the blades, reducing the bending moment and the
stresses in the vane due to this constraining effect. However, it is also
necessary to recognize that the coverband can also induce move-
ment or displacement in the vane radial position. Various effects can
cause this displacement:

256
Steam Path Component Alignment and Stage Spatial Requirements

Tenon hole in coverband misplacedThe production of the


tenon hole in a coverband (attached by riveting) is an opera-
tion that can be done on-site if the bands are not pre-
punched. This tenon hole production is undertaken using a
punch and die. Any error in locating the hole will require
the vane to be distorted by some small amount to facilitate
assembly. In small radial height blades it is unlikely the vane
can be adjusted, but in the longer elements this is entirely
possible

Vane pitchingThe vane tips on certain elements (in stages


using tangential entry blade roots) may have their pitches
removed from the design value to facilitate blade window
closure. If pre-punched coverbands are used, it is possible
the vanes be distorted to allow this assembly. Again, consid-
erations of the effect of the vane and its ability to be deflect-
ed allowing assembly must be considered

Incorrect tenon positionThe production of a tenon at the


blade tip is normally a separate manufacturing operation. If
some condition exists that will allow the production of a
rogue tenon on one blade element (due possibly to the inclu-
sion of other debris in a jig), this can cause error in the posi-
tioning of the tenon

Coverband distortion is not a common occurrence, and often


occurs when clearance between the tenon and coverband hole
assembly is achieved. However, it must be remembered that as the
tenon is formed by the riveting operation, there will be some degree
of vane distortion as the gap between the tenon and hole is closed.

Method of vane location. The method of locating the vane with-


in the steam path can influence the axial location. When vanes are
attached to a root through a platform, it is relatively easy to ensure
the correct axial position within design tolerances. If, however, the
vanes are located as part of a fabrication or casting process (as often

257
Turbine Steam Path Maintenance and RepairVolume One

occurs on many stationary row elements), then axial misplacement


is common. When establishing tolerances with these forms of con-
struction for stationary rows (the diaphragms), the designer recog-
nizes these vanes are of a form permitting adjustment after comple-
tion of the assembly or manufacturing process.

Quantifying acceptable axial placement errors is extremely


difficult. In addition to the requirements of achieving a throat that
will contribute to and control the discharge area, there are consider-
ations of the expansion passage divergence. In general, the greater
the turning angle , the more precise the axial pitching must
become. Consider the three profiles R, S, and T in Figure
2.14.2. With a large turning angle, as profile S moves downstream,
the throat at the entry to the expansion passage (formed between
profiles R and S) will close down at the inlet, causing a passage
that is divergent throughout its length. This is unacceptable.

Pitch error
The theoretical pitch at any diameter can be determined from the
diameter being considered and the number of blades in the row (See
equation on page 246). The actual pitch at any position on a blade
row will often differ from the theoretical value. It can be expected
and accepted if within tolerances.

The effects of pitch error on an individual cascade needs to be


considered in terms of their effect on the passage shape and the dis-
charge area from the individual throats. It is obvious the total pitch
around a row must be equal to the sum of the tangential pitches of
the individual blades. A plus error in one pitch must be countered
by a negative error in anotheror in a series of others.

Consider the three vanes R, S, and T in Figure 2.14.7. The


normal pitch Pn is shown for the three profiles. The design value
for profile S is shown in the design position. As a pitch error

258
Steam Path Component Alignment and Stage Spatial Requirements

occurs, the pitches modify to Pr-s = Pu, and Ps-t = Po. The sum
of these pitches remains the same as the original pitches. However,
the distribution has changed. In fact, in a complete blade row there
may be no pitches at the design value, but their mean must always
be equal to the design.

T S R

O
O

Po Pu
Pn Pn

Fig. 2.14.7Three vanes withFigure 2.14.7


the center vane off pitched.
Three vanes with the center vane off pitched.

The throats formed from discharge tail to adjacent suction face


will also vary with pitch. The form of the throat and the ratio O/P
will depend upon the shape of the suction face. It can be seen that
the point on the suction surface at which the throat originates pro-
vides a clear indication of the steam discharge angle. Consider the
two forms of discharge tail suction face profile that can be used
straight suction face and curved suction faceand their effect on the
discharge throat and angle:

Straight suction face. The suction face at the point of minimum


discharge width is formed on a straight surface, as shown in Figure
2.14.8. In this case the value of the discharge angle (o or 2) is
unaffected by pitch changes. There is, however, a small change in the

259
Turbine Steam Path Maintenance and RepairVolume One

ratio O/P, when the effect of the tail thickness b is taken into
account. However, the value of b is sufficiently small that change
can often be ignored.

Large pitches are the one exception to this constant O/P ratio.
In these cases, when the throat is formed of the radius R1 for the
value of pitch, the throat no longer follows the linear relationship.
This condition will often occur on constant profile vanes (cylindrical
profile) at tip sections where the pitch has increased sufficiently.

R1

Oo
u n
2
b Pu
n
o

Fig. 2.14.8The
Figureeffect of off pitch-
2.14.8
ing
Thewith a of off pitching with a
effect

Curved suction face. If the discharge portion of the suction face


is curved (Fig. 2.14.9), the throat is formed at a different discharge
angle, providing an angle varying from 2u to 2o as shown by
the tangent lines. Therefore, the effective throat and the ratio O/P
are functions of the pitch, for all values.

This form of vane is therefore more susceptible to the discharge


angle variation and is dependent upon the accuracy of manufactur-

260
Steam Path Component Alignment and Stage Spatial Requirements

ing pitch. It can also be seen by inspection that the discharge curva-
ture of the tail suction surface should be kept as large as possible
minimizing the rate of angle change.

There are a number of contributing factors causing pitch error.


These may occur individually or in combination. The most common
of these are listed here.

Oo
u n
b

2 Pu
Pn
u Po
n
o

Fig. 2.14.9The effect of off pitching


Figure 2.14.9 with a curved tail
blade. The effect of off pitching with a
curved tail blade.

Root block thickness-tangential entry roots. For blades that enter


the wheel in a substantially tangential direction, and are contained
on root blocks that are in face contact on the wheel, the tangential
pitch of the block will establish the pitch at the root section.
Therefore, if the vane and root are correctly aligned, the pitch will be
correct along the radial height.

261
Turbine Steam Path Maintenance and RepairVolume One

Root block thickness-radial entry roots. In addition to the tan-


gential entry blades, some radial entry pinned roots can have small
tangential clearances between the blocks, or alternately mate to
adjacent surfaces. There is the possibility that pitch errors can be
introduced if the pitches of the individual blocks are not correct, or
if having a small tangential pitch, the gap between the blocks is not
maintained at design values.

Root block wedge angle. For tangential entry blades, the vanes
are produced on a root platform, which has a plan shape suited to
joining the vane to the greatest extent possible over its entire perime-
ter. In general, the root platform of these tangential entry blades is
manufactured so there is adequate contact between adjacent sur-
faces when they are assembled to the wheel.

Consider a section through a root block (Fig. 2.14.10). Here the


pitch at the root platform face is shown as Pr, and at the root block
has a total depth Hx. As the distance from the outer surface
decreases, the effective pitch will reduce so that at the base of the
blade root diameter Da, the pitch has reduced to Pa. It is normal
for the block to have its back (suction) face parallel to a line G-G
that passes through the center of gravity of the vane profiles. To
achieve the correct wedge angle , the front (pressure) face must
converge to a reduced pitch at the base of the block.

The value of can be determined from the number of blades


in the row.
= 360
Zb

where:

Zb is the number of blades in the 360 of the row

It is important to consider the required accuracy of the wedge


angle to ensure the blade pitch is not adversely affected by any
errors. In assessing what this tolerance should be, it is necessary to

262
Steam Path Component Alignment and Stage Spatial Requirements

Dt

Pr

Hx Dr

Pa
G Da

Fig. 2.14.10A
Figuresection
2.14.10through a
blade vanethrough
A section and roota block
blade showing
vane and
theroot
alignment between
block showing thethe vane and
alignment
block.
between the vane and block.

consider the functions of the root block relating to blade spatial


requirements. There are four basic requirements to be considered:

The blade vane be held in the correct radial alignment rela-


tive to both adjacent elements, and the radial line describing
its center of gravity passes through the center of the rotor

The root block be held at the wheel or rotor position so that it


is unable to make any movement in the tangential direction.

263
Turbine Steam Path Maintenance and RepairVolume One

Such movement occurs when there is inadequate contact


between the root block facesthere is contact at one diame-
tral position only

The root is unable to rotate in the root slot by unacceptable


amounts

The root block faces make contact eliminating vibration and


reducing the possibility of fretting corrosion in the interfaces
positions

In addition to these placement requirements, it is also essential


the root block be placed so the centrifugal and other loads devel-
oped in the vane during operation are transferred to the rotor even-
ly over any number of load bearing surfaces.

To achieve these requirements it is necessary for the root block


angle be held at the +/- 0.10 level. This can appear to be exces-
sively tight, but it represents a difference in the values of Pa and
Pr of only +/- 0.0017" per inch of block height between the two.
Therefore, if the pitches Pr and Pa are specified to a tolerance of
design value of +/- 0.001"a normal tolerance for blade pitchthen
the root angle should be acceptable.

As an example, consider the blade root block shown in Figure


2.14.11, which is one of 120 blades in a row. The design diameters
are shown as Da=32.14" and Dr=36.84." The blade root block is
therefore 2.35" high.

At the extremes of tolerance, the wedge angle will vary


between:

-1 0.9655 - 0.8404 = 3.047


max.
= tan
2.35
-1 0.935 - 0.8424 = 2.950
max.
= tan
2.35

264
Steam Path Component Alignment and Stage Spatial Requirements

The actual permissible angular variation in is from design


3.047 to 2.950. These values are well within the engineering tol-
erance band of +/- 0.1 from the 3.0 design value.

Pr = 0.9654"
+/- 0.001"
G G

Dr = 36.84"

Hx = 2.350"

G G

Da = 32.14" Zb = 120 pitches


Pa = 0.8414"
= 3.00
+/- 0.001"

Fig. 2.14.11The root geometry and block


Figure form.
2.14.11
The root geometry and block form.

Setting angle. The setting angle of the vane can have minor
effects on the pitch between blades. This is not a major factor, but if
there is a variation of setting angles between adjacent elements
because of the method of manufacture, their pitch can be affected.

Vane position on the root platform. The vane can be misplaced


on the root platform. This is not common, but for certain methods of
manufacture there can be errors effecting tangential placement.

265
Turbine Steam Path Maintenance and RepairVolume One

Vane lean. In addition to having the possibility of being misplaced


on the root platform, the vane can also be produced with a lean. This
lean can occur in either the tangential (pitch) or axial direction.

Fig. 2.14.12The effect of vane lean on blade S. The variation of


tip pitch can be seen as error dP.

The effect of this tangential lean is illustrated in Figure 2.14.12.


Here, three blade vanes (R, S, and T) are shown in their final
positions as mounted on a rotor. Also shown are two additional blades
(J and K) adjacent to this group and pitched correctly. Of this

266
Steam Path Component Alignment and Stage Spatial Requirements

three-blade group, the two outer vanes (R and T) have their vanes
in a correct radial alignment (G-G). However, the center vane (S)
has been produced with a tangential lean towards T by an amount
dP at the tip, as shown. This lean produces tip pitches of Pr-Ps=Po
and Ps-Pt=Pu. The effect of this off pitching is shown in Figures
2.14.7 and 2.14.8. Figure 2.14.7 shows the effect on discharge throat
and angle for vanes having a curved discharge suction face. In addi-
tion to the effect on passage shape, additional bending moments are
introduced and the consequential stresses into the vane.

When it is determined that vanes are out of radial alignment after


assembly, considerations of this error need to be reviewed. These
considerations may be of no consequence, however, they can have
the potential to impact on the performance of the unit and require
detailed evaluation. These consideration include the following:

Pitch error vs. height. The vane can be off-pitched at one radial
location but may also be bent. The implications of this are that the
pitch can be within tolerance over a major portion of its radial length
and then be outside over the remaining portion. Therefore, when an
off-pitch condition is found, it could be advisable to consider this
possibility, particularly with longer blades formed by manufacturing
methods that induce bending and twisting, or blades that have
employed a field repair involving the use of heat, which causes some
forms of distortion.

The possible effect of the coverband. A coverband can cause


pitch distortion, which can be localized over a portion of the total
length. This condition occurs when a blade has tip distortion and the
vane is bent to allow assembly. This is a complex issue, as the vane
will have bending stress induced in it to facilitate assembly.
However, the vane would also be stressed by the off-pitch condition
and possibly in operation the bending stress could be reduced as the
vane attempts to attain a true radial position. There is normally suffi-
cient margin in the factor of safety that this should not be a concern.

267
Turbine Steam Path Maintenance and RepairVolume One

Vane untwisting if straightened. If a vortex vane is found to be


bent in the pitch direction, the effect of straightening it may also
cause the vane to untwist and affect the setting angle and expansion
passage form, as the vane will bend in the tangential direction by
untwisting about its axis. Therefore, bending to achieve coverband
assembly should be limited by the amount of untwist this introduces.

In an actual design, the designer is aware of the inevitability of


vanes being off-pitched, and allows for a margin in both stresses and
efficiency guarantee. However, there are limits placed on this effect
and should elements be outside these after assembly, then it is nec-
essary for the designer to examine and rule on their acceptability or
define the corrective action required.

Method of manufacture (casting and welding). The methods of


manufacture can have a considerable effect on the final pitch
between the various elements, as well as the methods of correction
available to conditions that exceed design tolerances. The rotating
blades are normally manufactured with considerable care, and the
possibility of errors after assembly is minimized to the greatest extent
possible. Stationary elements manufactured for direct assembly to a
blade carrier or inner casingand manufactured to the same dimen-
sional criteria as rotating blade elementstend to meet design crite-
ria within close tolerances. However, those stationary elements
particularly for diaphragm stages constructed and attached to inner
and outer bandsoften employ a casting or weld fabrication
process. These methods tend to introduce greater variation in the
pitch, but they also allow a greater degree of corrective measures
after completion of the primary manufacturing process.

The steam path dimensional audit


At completion of manufacture and assembly to the rotor or cas-
ing, it is often useful to undertake an audit of the major blade dimen-

268
Steam Path Component Alignment and Stage Spatial Requirements

sional values that have the potential to affect the performance (effi-
ciency and structural reliability) of the unit.

Figures 2.14.13 (a), (b), and (d) show the measured values of
pitch and throat at the tip and mean diameters and radial discharge
height of 14 elements in a blade row, having a design radial height
of 12.08". In (a), the measured values from the tip section of these
blades are shown.

Tip
0.59 diameter
0.58 1.8
Throat 'O'

0.57

Pitch 'P'
Dt 1.7
0.56
Blade radial Height 12.06"

0.55

0.54 1.6
(a) Blade batch end points

Dm Mean
0.53 1.6
diameter
Throat 'O'

Pitch 'P'
0.52
0.51 1.5

0.50
(b) 1.4
Dr
Mean
+5% diameter
'O/P'
0
(c)
Tip
-5%
diameter

1 2 3 4 5 6 7 8 9 10 11 12 13 14

12.10
Height 'H'

12.08
12.06
(d) 1 Blade group
Discharge

6.5
area 'Ad'

6.3
6.1
(e)

Fig. 2.14.13The results of a partial audit Figureon2.14.13


14 throats formed between 15 blades in
The results of a partial audit on 14 throats formed between 15 blades in a rotating blade row.
a rotating blade row.

They are grouped by a coverband connecting blades and having


coverband discontinuities, batch end points, at throat 4 and 11, the
throat and pitch are shown. In (c), the ratio O/P for this tip section

269
Turbine Steam Path Maintenance and RepairVolume One

t is shown. It can be seen that there is considerable variation of this


ratio well outside what might be considered an acceptable range.
Similar values of throat and pitch are shown in (b) for the mean
diameter Dm, and again in (c) the ratio O/P. These values are
more realistic, indicating these blades may have become bent or
twisted in their outer sections.

Fig. 2.14.14Measured values of blade axial position relative to a


mean inlet and discharge edge.

270
Steam Path Component Alignment and Stage Spatial Requirements

A similar audit can be made of the inlet and discharge edges,


showing the possibility of proud and recessed blades. This effect
will exist to a certain extent on all blade rows, but is normally sig-
nificant only on those longer elements where the ability to, and
probability of distortion exists. Figure 2.14.14 shows the measured
values of ten 17.27" blade elements in a freestanding row. These
measurements were taken at the inlet and discharge edges at a dis-
tance about 1.0" below the tip. In a normal audit, it is not known
where the actual edges should be located; therefore, the measure-
ments of +/-da are relative to an undefined edge. Such an audit
can give unquantified information on the quality of blade production
but the manufacturer normally has standards for da, which should
be met.

Such audits are not expensive to conduct, and are required only
if a visual inspection suggests problems and excessive variations
from design tolerances are present. In the case of a rotating blade
with an attached erosion shield, it is possible some deformation will
occur as a consequence of the heating process as well as the actual
position of the shield on the vane.

Setting angle error


The setting angle shown in Figure 2.13.3 is a critical charac-
teristic of the row, as it defines the form of the expansion passage and
therefore the potential efficiency of the row. Unfortunately, the set-
ting angle is a characteristic which is relatively difficult to gauge
because of possible small errors that may occur in form at the inlet
nose and discharge tail. It is at these locations where relatively small
changes of form can indicate an error in the setting angle of a vane
which is, in fact, properly set. However, there are errors within the
individual stage components that can occur and do affect the setting
angle producing an expansion passage that may not meet design
requirements within tolerance.

271
Turbine Steam Path Maintenance and RepairVolume One

Many of the types of errors discussed previously also affect the


setting angle and introduce errors exceeding the design require-
ments. There are also other considerations, which include the fol-
lowing:

Root block twist. There are root forms that are entered into, and
which during operation are surrounded by the wheel. These internal
root forms are of a generic class often called the inverted T root.
These roots have a central stem that has some small clearance from
the stem to the wheel groove sides (Fig. 2.14.15).

The design specification would indicate there is equal clearance


on both sides of this stem, and that this is maintained during opera-
tion. In fact, when the root block form is a lozenge, there is a ten-
dency for the root platform to rotate and tilt the blades in the groove
(Fig. 2.14.15). This twisting has an effect on the total spatial rela-
tionship of the cascade formed between the vanes, affecting the set-
ting angle , the ratio O/P, the discharge area Ad, the row
width, and many other characteristics of the row.

P Wheel rim d


Blade
root

Fig. 2.14.15Showing the effect ofFigure


a lozenge root untwisting in the wheel due to gap
2.14.15
d. This untwist
Showing thewill rotate
effect of aalozenge
blade through an anglein.
root untwisting the wheel due to gap d.
This untwist will rotate a blade through an angle .

Blade untwist. The effects of blade untwist were considered pre-


viously, and the possible effect on pitch is shown in Figure 2.14.6
(assuming the vane were to rotate about its center of gravity G).

272
Steam Path Component Alignment and Stage Spatial Requirements

However, no consideration was given to the effect of this untwist on


the setting angle and expansion passage form.

Fig. 2.14.16The various forms of error which can occur in a cascade. Showing in fig-
ure 10.6.16(a) is the design condition.

Figure 2.14.6 considers one such profile. However, when exam-


ining the effects on setting angle, it is necessary to consider the effect
of multiple blades with untwists, and if the untwist is caused by some
phenomenon such as a heating process. In each case, there is no rea-
son to believe each blade will untwist to the same degree. The effect
of untwist on the setting angle is somewhat more complex.

273
Turbine Steam Path Maintenance and RepairVolume One

Conditions can only be considered in general. Each condition will


tend to be different and dependent upon the untwist of two vanes.

In terms of expansion passage form, convergence rate, design-


setting angle, and discharge area and angle, it is necessary to con-
sider the point on the profile about which the vane has rotated. This
point of rotation can affect all of the above parameters. Figure
2.14.16(a) shows the design-specified condition of profile J and its
position relative to the discharge tail of profile K. These profiles
have center of gravity position G at a distance An and Tn from
the discharge point on the profile. For this condition, the form of the
expansion passage is shown in Figure 2.13.13, which is convergent
from inlet to discharge with a final throat O shown as On in
Figure 2.14.16.

If the vane J rotated about some point within the cascade, there
are an infinite number of variations of the major parameters that can
be achieved, due to the position of both the twist in profile J and
also movement in profile K. Consider the following three cases, in
which considerations are simplified by assuming profile K remains
in its design specified position:

Profile J rotates about the centroid G. In this situation


[Fig. 2.14.16 (b)], there will either be an increase or decrease
in the setting angle by d. In either case, the form of the
expansion passage will modify, causing a change in the pitch
and throat

Profile J rotates about the discharge tail. In this case [Fig.


2.14.16 (c)], there will be a change in the throat, but these
changes will have a smaller effect on the throat value. The
width will either increase or decrease by an amount da,
depending upon the direction and extent of twist

Profile J rotates about the inlet nose. In this change, rota-


tion occurs about the inlet nose [Fig. 2.14.16 (d)], the conse-

274
Steam Path Component Alignment and Stage Spatial Requirements

quence of which can be an exaggerated change in the throats


for any degree of twist. Again, the width will change by
amounts +/-da, depending upon the direction of twist

In establishing the extent to which twist is acceptable, the


designer should avoid any condition which does not ensure that the
expansion passage remains convergent from inlet to discharge (un-
less a converging/diverging form is required by design). The extent to
which twist is acceptable is established by the design specification.

The profile shown in Figure 2.14.17 is for a low-reaction vane,


normally defined as an impulse design having a small degree of
reaction at its root section. In this design, the ratio in the inlet throat
Oe to discharge On is equal to 1.40. In such a cascade, any rel-
atively small change in the setting angle has the potential to modify
the passage shape and convergence to the extent there is a consid-
erable loss in stage efficiency.

Fig. 2.14.17The divergence ratio in a


blade row.

Vane lean. The effect of blade vane lean was considered earlier
in terms of its possible effect on axial and tangential (pitch) distor-

275
Turbine Steam Path Maintenance and RepairVolume One

tion. It is also possible for lean in the axial and tangential direction
to have an effect on passage form. In general the greater the vane
turning angle , the less it is able to accept lean in either direction.

Axial directionIn the axial direction, the lean will modify the
discharge throat and, therefore, the area. If a rogue blade exists
in a total row, the increase in throat of one passage is coun-
tered by the closing of another. However, the most common
situation is a majority of blades in a row proud or recessed
from the theoretical discharge edge (Fig. 2.13.12)

Tangential directionLean in the tangential direction will


produce a condition in which the passage throat will
increase or decrease, depending upon the direction of lean.
This form of error can distort the form of the expansion pas-
sage to the extent a diverging/converging condition could
exist. The greater the vane turning angleand the throat inlet
to discharge ratio approaches 1.0 (Fig. 2.14.12)the greater
this extent can be

These three forms of erroraxial placement, pitch, (refer to pre-


vious text concerning pitch error), and setting anglehave been
shown to affect many important characteristics of the cascade blade
row. It is not possible to manufacture a blade row eliminating these
errors at an acceptable cost. The design function recognizes this, and
establishes manufacturing and assembly tolerances within which
variations are acceptable. After assembly, if these requirements are
not met, it is the responsibility of the design function to evaluate and
rule on acceptability of corrective action. Corrections that are made
must be fully evaluated, and qualified personnel must undertake the
evaluation process.

276
Steam Path Component Alignment and Stage Spatial Requirements

The discharge area and angle


The discharge area and angle are determined in terms of the
physical dimensions of the vane and their setting at discharge from
the expansion passages. These characteristics of the rows must meet
design specification or the performance of the unit will deteriorate.

The audit is a meaningful method of checking compliance, but


unfortunately, in rotating elements it is not possible to take correc-
tive action. Therefore, the level of compliance is dependent upon the
expertise and effort taken to manufacture the individual blade ele-
ments. This also applies to stationary blades that are inserted in the
blade carrier or casing. For stationary elements, adjustment is often
possible (see chapter 7).

The swallowing capacity


The swallowing capacity of a blade passage was previously
defined as the quantity of steam that would enter a specific passage,
and in a blade row that was perfectly spaced, each passage would
carry an equal quantity of the working fluid.

Should there be any misplacement of vanes within the cascade


that affects the passage inlets, the quantity of entering steam will be
modified. Also, in any row extending for 360, any error in one pro-
file will modify the swallowing capacity of two expansion passages.

Consider the profiles R, S, and T in Figure 2.14.18, in which


profile S is misplaced. The steam enters the row at a relative angle
1. This profile S can be misplaced by any of the conditions
shown in Figure 2.14.16. Because the steam enters the row in what
might be termed a division by opportunity, the steam will divide at
the inlet nose, with the largest quantity flowing into the largest inlet
throat. Therefore, if profile S were proud by an amount +da (Fig.
2.14.18), the passage formed between profiles R-S would receive

277
Turbine Steam Path Maintenance and RepairVolume One

more steam than that formed by profiles S-T. The quantities would
be reversed if S were recessed by a similar amount.

Similarly, a pitch error +/-dp will influence the steam quantity


entering the individual passages. Setting angle error will also have an
effectthe magnitude being a function of the degree of error and the
point in the profile about which the vane has twisted. If the profile
shown in Figure 2.14.19 is considered (with the inlet angle 1
shown as 90), then the influence of proud and recessed blades is
not a consideration. However, the pitch and setting angle will still
influence the flowing quantity, as discussed, for the blade with a
larger turning angle (Fig. 2.14.18).

Fig. 2.14.18Three blades showing the effect of a misplaced profile on


swallowing capacity, for a vane with a large turning angle.

There are no general rules that can be developed concerning


potential detrimental effects of swallowing capacity variation from
design, and each situation should be considered separately. Suffice it
to say, variation in the steam quantity entering a passage will influence
the performance of a row, and there is a limit set by design establish-
ing the tolerances within which variation is acceptable.

278
Steam Path Component Alignment and Stage Spatial Requirements

Fig. 2.14.19Three blades showing the effect of a misplaced profile on swal-


lowing capacity for a typical stationary vane.

Consider the profiles in Figure 2.14.20, in which profile K is


misplaced as shown (chain line) relative to profile J. This mis-
placement can be a combination of errorspitch dp, axial setting
da, and setting angle d. These are sufficient to make inlet areas
different. With the misplacement shown, area Ajk between profiles
J and K is larger than area Akl between profile K and L.
Therefore, passage J-K has the potential to admit a larger steam
quantity. However, the discharge area Ajk is smaller. This total
effect is to cause considerable variation of pressure and velocity dis-
tribution in the various passages.

There are several possible consequences on row performance


associated with a variation in the individual passage-swallowing
capacity that should be considered.

Steam discharge velocity. The quantity of steam discharging


from passage J-K is different to that discharging from K-L. If the
discharge areas from both throats were the same, then the velocity
from J-K would be larger. However, in addition to the larger flow
quantity (Fig. 6.14.20), the throat of J-K has also decreased relative

279
Turbine Steam Path Maintenance and RepairVolume One

to K-L. This reduction in area will also cause an increase in the


velocity from J-K relative to that from K-L.

Fig. 2.14.20A misplaced vane with the resulting stage parameters and
inlet flow area.

Steam discharge relative angle. The value of the effective dis-


charge angle is a function of the ratio O/P. It can be seen that both
pitch and throat can vary from the design values, as a consequence
of misplacement error, and therefore the discharge angle will also
change.

In the situation shown in Figure 2.14.20, the passage J-K will


have a larger discharge velocity (Cjk) at possibly a smaller dis-
charge angle than the design. Also, passage K-L will have a lower
discharge velocity (Ckl) at a larger angle. The effect of this on
steam entering the following row is shown in the velocity triangle for
row Y in Figure 2.14.21, where the effect of both velocity and inci-
dence in entering row Z can be clearly seen.

280
Steam Path Component Alignment and Stage Spatial Requirements

Steam bending stress. The steam-bending stress in a blade is


directly proportional to the quantity of steam flowing across it, and
the angle through which it is turned. Therefore, the stress induced in

Fig. 2.14.21The velocity triangles for expansion passages J-K and K-


L of figure 2.14.20.

these blades will not be the same, because the flow quantity varies
from one expansion passage to another, and discharge angles are also
different from passage to passage. The effect of the discharge angle is
relatively small compared to the effect of the changing flow quantity.

281
Turbine Steam Path Maintenance and RepairVolume One

REFERENCES
1. Glatt, P., Laser Measures Centerlines Quickly, Economically,
Turbomachinery International, January/February, 1994

2. Jordan, S., and M. J. Fraser, Design Modifications and Repairs


to Existing Steam Path Components to Improve their Existing
In-Service Performance, EPRI

3. Steam and Combustion Turbine Blading Conference,


Orlando, FL, January 1992

4. Hollingworth, K., The Application of Weld Repair Techniques


to Reduce Costs and Outage Time on Steam Turbines, Turbo-
machinery Maintenance Congress, Singapore, 1988

5. Rasmussen, D.M., and W.T. Durbin, Steam Turbine Case


Repairs to Extend the Operating Life, ASME Paper 84-JPGC-
PWR-44

6. Sanders, W.P., and I. Schulz, Steam Turbine Rotor Restora-


tion, Eighth Turbomachinery Maintenance Congress, Kuala
Lumpur, Malaysia, October 1992

7. Jones, G.T., and J. Gunning, Causes for and Methods of


Straightening Bent Turbine Rotors, Eighth Turbomachinery
Maintenance Congress, Kuala Lumpur, Malaysia, October
1992

8. Phonic, T.R., Turbine Maintenance and the Straightening of


Shafts, Journal Of South African Institution of Mechanical
Engineers, February 1953

282
Chapter

3
Steam Path Damage
Induced by Water

INTRODUCTION
The steam that enters the turbine steam path has a high-energy
content. In fossil-fueled cycles, this steam is normally superheated,
i.e., the steam contains a degree of heat in excess of that required to
completely evaporate it. On expansion, this superheat is released in
the steam path, converting its thermal potential energy to kinetic
energy in the rotor. When all the superheat has been converted to
kinetic energy, further expansion causes the remaining dry, saturated
steam to give up a portion of its latent heat, which is converted to
water. The flow then becomes a two-phase mixture of steam and
water.

However, in the water-cooled nuclear cycle and in the majority


of geothermal cycles, steam that enters the high-pressure section
contains some small amount of moisture. This moisture portion of

283
Turbine Steam Path Maintenance and RepairVolume One

the mixture is increased throughout the initial stages of the expan-


sion and the majority of the low-pressure stages.

Once water has formed, it must be removed or transported


through the steam path by the parent steam. Because of the relative-
ly larger mass of the moisture particles, they are unable to pass
through the blade passages with the same degree of acceleration as
the dry steam. These water particles must have work done upon them
in an attempt to match steam velocities.

These steam velocities will act upon the water particles because
the steam accelerates and changes direction in the blade passages.
However, the water particles will be unable to accelerate and
achieve velocity equality because of their larger mass. Therefore, the
water particles will have trajectories within the steam path that are
different from those of the steam. Many will collide with the surfaces
of the steam-path components, upon which they will be deposited.
After deposition, this moisture will be unable to re-enter the main
steam flow, but will still experience large field forces from the flow-
ing steam. This tends to make the resulting water film flow in the gen-
eral direction of the steam particles. With moisture particles con-
stantly being deposited on the surface from continued contacts,
there will ultimately be a significant film of water flowing across the
steam path surfaces.

It is this deposited moisture that has the potential to cause various


forms of damage to the steam path components as it collects into
pools and re-enters steam flow as larger droplets. In addition, that por-
tion of the transported moisture remaining in the steam will reduce
steam path efficiency, though its presence may not contribute to any
mechanical damage to the steam path components themselves.

There are sources of water other than those formed by conden-


sationthree principle sources of water within the steam path alone.
The first two listed here are water-contained in the steam path, but
restricted to existence in the saturated region. The entire list includes:

284
Steam Path Damage Induced by Water

water formed by condensation of the working fluid during its


expansion through the stationary and rotating blade rows

water sprayed into the low-pressure section exhaust to


remove excess heat generated by frictional heating during
part or light load operation (this will be considered in greater
detail)

any portion of the unit connected to other equipment within


the steam cycle, i.e., water re-entering the steam path from
external sources

WATER CONDENSATION
IN EXPANDING STEAM
As the steam expands, it first releases its superheat energy until it
reaches the saturated condition. Then, with further expansion, a por-
tion of the latent heat contained in the steam is released. This con-
version of latent heat introduces a state where water is formed in the
expanding steam. This water exists as fine particles (fog) that are
transported by the parent steam through the steam path until it
exhausts to the condenser, or impacts with a surface to which it
adheres.

Consider such an expansion on the Mollier Diagram (Fig. 3.2.1).


The steam expands from condition A to condition D, at which
point the steam is dry and saturated. It would be expected from the
previous discussion that with further expansion, water would form in
and be transported by the steam. However, for heat to transfer from
the gaseous to liquid phase requires a finite period of time, and the
expansion of steam in the steam path is extremely rapid.

285
Turbine Steam Path Maintenance and RepairVolume One

Note: The elapsed time for steam entering a high-pressure sec-


tion to expand through it, and through the reheat and low-pressure
sections is about 0.2 seconds, if the time in the crossover pipes and
the boiler reheater section is ignored.

A
Sat. X
= 1.0 Supersaturated
D condition
from "D-S"
Pi
Pn
Fog Xi
seeds Wilson
form Line at Xi
Ps Pn

Xi
Ps
ds

Detail of entropy
increase "ds" at heat
transfer from steam
P to water phase
e

Fig. 3.2.1Showing supersaturation of


steam at the saturation line.

Because heat transfer cannot occur instantaneously, the expan-


sion will continue as shown in Figure 3.2.1. Here, expanding steam
crosses the saturation line at D, and expands into the moisture
region, achieving a super-saturated condition such as S. At that
point however, the transfer of heat will have been completed and the
mixture will reach, or approach thermal equilibrium conditions, and
moisture will form. This is the point at which a fog is formed, con-
sisting of particles from about 0.5 to 1.0 microns in diameter. It is
upon these seed particles that further deposition will occur in their
passage through the steam path.

286
Steam Path Damage Induced by Water

Once moisture particles have formed, various forces developed


by the steam will influence them. This includes the drag force that
causes the moisture particle to attempt to follow the steam flow
velocity, being both accelerated and retarded through the blade pas-
sages. Because of their relatively greater mass, these moisture parti-
cles will not be able to follow the steam velocity, and will continue
to extract energy from the steam, in an effort to achieve and main-
tain steam velocity. In addition, because of their greater relative
mass, the water particles are unable to change direction as effective-
ly as the steam, and will therefore assume a different trajectory or
path in their flow between the blade passages.

The manner of water formation


There are three possible means by which water may form in the
steam path:

Condensation on microscopic particles often impurities or


contaminants contained in the working fluid. However, mod-
ern methods of feed water treatment would preclude the
contamination to any excessive degree. While there may be
some impurities present, attracting a little condensation, the
quantity is not likely to be sufficient to form a significant por-
tion of the total moisture present at any time. Some of the
impurities will combine with water, and this has other more
insidious implications in terms of corrosion, but is inconse-
quential in terms of moisture formation and flow.

Condensation on the surfaces of the steam path components.


For steam particles present in the parent steam to condense
on the surfaces of the steam path, they must contact it and be
at a temperature that will cause condensation at the local
ambient pressure. However, these metal surfaces are always
covered by their boundary layer, which tends to be at a mar-
ginally higher temperature than the surrounding steam. This

287
Turbine Steam Path Maintenance and RepairVolume One

is because the velocity energy of the steam in the boundary


layer has been reconverted back to thermal energy, raising
the temperature of the steam. Therefore, the surfaces that the
moisture particles contact are above the temperature of the
surrounding steam, and condensation is unlikely.

Spontaneous nucleation after a limiting degree of super-satu-


ration has been reached, and there has been sufficient time
for heat transfer. It has been shown by analysis that the only
meaningful manner in which sufficient moisture can form in
the steam is by this process.

It can be seen from these considerations that the moisture pres-


ent in the steam path is essentially a consequence of spontaneous
nucleation, and the formation of minute water droplets (which for
convenience will be referred to as the water seeds) can be shown to
have a size defined by the Kelvin-Helmholtz:

Ln { }
p
po
=
2.
RTsr

where:

p = Steam pressure
po = Vapor pressure corresponding to seed temperature Ts
= Surface tension
= Seed density
r = Critical radius
R = Characteristic gas constant

Once these seeds have formed, they will continue to grow as


expansion continues and further moisture is added. It is debatable
whether droplets coagulate during their passage through the steam
path, and it is not clear if smaller droplets (seeds) merge into larger
drops (reference following sections on Measuring moisture

288
Steam Path Damage Induced by Water

Distribution and Content and Water Removal from the Steam Path).
However, other mechanisms for collection, present within the steam
path, allow these droplets to agglomerate to much larger and more
destructive forms.

Upon the formation of the water seeds, there is a small rise in the
local pressure. This is shown in Figure 3.2.1 as an increase from Ps
to Pn.

RADIAL DISTRIBUTION
OF MOISTURE
In the previous section the phenomenon of super-saturation and
the delay in heat transfer from the gaseous to liquid phases was dis-
cussed. It is clear that if we consider the steam path of a low-pres-
sure section in the region of the saturation line, then the point of
moisture formation could be as shown in Figure 3.3.1. There would
be a small axial distance after the saturation condition that would
represent the under-cooled region of the steam path. This distribution
assumes the pressure is constant at all radial positions along the axial
length of the steam path.

Entry to the under-cooled region is pressure related, and origi-


nates at the point where the expansion crosses the saturation line.
However, because the pressure gradient in the blade path is not radi-
ally constant at all axial positions, the position of the under-cooled
region will not occur at a radially constant axial position. Figure
3.3.2 shows the measured pressure gradient in a low-pressure
expansion. If the position of the onset of under-cooling is related to
pressure, then this region will be of the form shown in Figure 3.3.3,
with the moisture, as a fog forming as the heat transfer is completed
at some lower pressure.

289
Turbine Steam Path Maintenance and RepairVolume One

Figure 3.3.1The position of the supercooled


region with no radial pressure gradient.

Figure 3.3.2The pressure gradient in a low pressure stage.

290
Steam Path Damage Induced by Water

Figure 3.3.3The pressure gradient in a stage with measured


pressure distribution.

The axial position and entry to the under-cooled region can be


established by measurement and calculation. However, it must be
remembered that any measured condition applies to one set of steam
conditions and flow quantities in any one unit. The position and form
of the under-cooled region is influenced by the quality of steam
delivered to the turbine from the boiler or reheater, the quantity of
steam flow, and (for any design) the stage heat drops and vane con-
figuration. Therefore, the regions shown in Figure 3.3.3 are steam
condition and load sensitive. Actual measured values of moisture in
the radial direction are shown in Figure 3.3.4.

291
Turbine Steam Path Maintenance and RepairVolume One

Figure 3.3.4Measured moisture distribution.

MOISTURE DEPOSITION
Moisture droplets formed in the steam are unable to follow the
direction of flow of the steam because of their relatively large mass.
Therefore, the moisture droplets will be unable to change direction
through the blade passages and will collide with, and be deposited
upon the various elements of the blade rows. There are two main
deposition areas to consider:

Deposition on the blade vanes


The path of dry steam flow between the stationary blade vanes is
shown in its basic form in Figure 3.4.1. This illustrates the steam flow
trajectories between a pair of stationary vanes. Shown are the stream-
lines and stream tubes of the steam flow between the pressure face of

292
Steam Path Damage Induced by Water

one profile and the suction face of an adjacent element. This station-
ary row is designed for a pressure drop to occur across it and for the
steam to be accelerated at the expense of this pressure/enthalpy drop.
Therefore, at discharge from the row, the velocity at L will be high-
er than at the inlet K, and the pressure lower.

Note: The concept of streamlines is used to aid in the explanation


of flow, and to represent a convenient means of anticipating the actu-
al conditions that can exist at any position within a flow passage.

Figure 3.4.1Steam flow lines through a fixed blade passage.

Now consider a moisture droplet at G, shown in Figure 3.4.2,


located on the streamline G-G. As the steam enters the influence
of the passage formed between the blade pair, the steam and the
moisture particle are accelerated through it. However, because of its
greater mass, the moisture particle is unable to accelerate at the
same rate as the steam, and it takes a path as shown, G-M (posi-
tion M being at the discharge point on the profile). The droplet will
ultimately collide with the pressure face and, after collision, will

293
Turbine Steam Path Maintenance and RepairVolume One

remain on this face, unable to re-enter the main steam flow, and
existing in the form of the original small particle. It can be seen, by
examination, that all moisture existing from the profile inlet nose and
position G, is being transported by the steam, and will be deposit-
ed on the profile surfaces.

Figure 3.4.2The trajectory of the moisture particle G which will collide with
the discharge tail.

Due to the forces exerted by the flowing steam on the deposited


moisture, this moisture will form a film that will flow across the pro-
file in a substantially axial (downstream) direction, to the point M
at the discharge tail. From this description, it is clear the quantity of
water that exists as a film on the profile pressure face will increase
as the distance from the inlet nose increases, and this is at maximum
at the discharge point (or discharge tail) of the profile (i.e., at posi-
tion M).

294
Steam Path Damage Induced by Water

Before considering the water path at the discharge tail, consider


the steam and moisture particle mixture that enters directly ahead of
the profile inlet nose Ni (Fig. 3.4.3). Here, as the steam and any
moisture particles it carries approaches the inlet nose, they must
divide to flow over either the suction face Fs or the pressure face
Fn. The division is a matter of which side of the profile exerts the
most influence on each of the individual steam and moisture parti-
cles. On the suction side of the profile, a portion of those moisture
particles deflected around that side will be unable to accelerate suf-
ficiently and will collide with the inlet nose as shown at position T.
This deposited moisture will then flow around the suction face,
influenced by the same forces that act on the pressure face.
However, the quantity on the suction face will not increase during
the flow across this area, as there is little or no possibility of those
moisture particles accelerating towards the pressure face impacting
there. On the pressure face however, the moisture particles that are
unable to avoid the inlet nose are added to those deposited (Fig.
3.4.2). These two sources represent the total moisture on the pressure
face Fn.

Figure 3.4.3Flow division at the profile


nose.

295
Turbine Steam Path Maintenance and RepairVolume One

At the position on the inlet nose where the flow begins to divide,
there is a slowing of the steam velocity as impact with the nose is
partially avoided. At this point on the vane profile there is a general
lowering of steam velocity, and with a lowering of velocity there is a
small local increase in steam pressure, (velocity energy is reconvert-
ed to pressure energy). This condition creates the stagnation point
S. The Mollier Diagram for this position is offered (Fig. 3.4.4) with
a steam inlet condition of A on the expansion line A1-A2. It will
be seen that accompanying the pressure increase dp (enthalpy
increase dH) is a small increase in the local temperature dT (1 to
1.5F). In fact, the inlet nose of the vane will be heated to some small
degree by this local increase in steam temperature.

Figure 3.4.4The steam conditions at


row inlet.

At the discharge point (shown as M in Fig. 3.4.2), the moisture


will accumulate as the quantity flowing down the pressure face con-
tinues to collect. However, the moisture cannot remain at this loca-
tion because of the force exerted on it by the high velocity of the
flowing steam. Consider the discharge tail (Fig. 3.4.5). The steam
expands through the passage created between the profiles, and will

296
Steam Path Damage Induced by Water

eventually discharge from the row at the throat (minimum area),


shown as O-O, with a velocity Cs. At this point the steam sub-
stantially leaves the influence of the profiles. At discharge from the
throat, the boundary layer on the suction face will continue to
adhere to the profile until some point is reached at which there will
be separation; this is shown as occurring at point Se. The void
caused by this separation will shed vortices, which are carried over
with the main steam flow into the following rotating blade row. Most
significant in terms of the moisture film and accumulating moisture,
is that the region of vortices is a low-pressure region, relative to that
existing in the vane discharge region. This means that the moisture
collected (or is collecting at the discharge point M of Fig. 3.4.2)
will be drawn around the tail, where it will accumulate with the
small quantity of moisture flowing down the suction face (Fs of Fig.
3.4.3) to form small puddles (or pools), Q.

Figure 3.4.5Moisture flow around a discharge tail, and collection on the suction
face.

This moisture accumulation puddle Q will continue to grow in


the vortex low-pressure region until it becomes too large to resist the
drag forces of the flowing steam. These puddles will then be torn
from the suction tail surface and re-enter the main steam flow as
considerably larger drops. Because the drops torn from the discharge
tail are no longer of a size that can be transported by the main steam

297
Turbine Steam Path Maintenance and RepairVolume One

flow, they break into many small droplets, but remain hundreds of
times larger than the original seed droplets, and are of sufficient size
to cause moisture impact erosion on the rotating blade inlet edge.

The manner of deposition in a rotating blade row is essentially


the same as in the stationary blade row, with certain differences in
the pattern of accumulation and flow. The physical considerations
affecting these differences and causing the flow pattern to alter are
listed below:

Within the rotating blade passage there are large centrifugal


forces acting on both the steam and moisture film. This cen-
trifugal force causes the deposited moisture to have a large
radial flow component on the surfaces of the vane

The passage shape may be designed for very low degrees of


reaction, and therefore the pressure drop may be lower. In
nuclear high- pressure stages (the normal impulse design) the
passage shape has a relatively small degree of convergence,
and the degree of reaction is low

In the rotating blade of an impulse stage, the velocity tends


to decrease rather than increase through the passage

Deposition on the sidewalls


Expansion through the steam passages results in moisture depo-
sition within them. We have discussed the deposition on the vanes,
however, in addition to water accumulation that occurs on the
vanes, there can be additional deposition on the sidewalls.

Consider the stationary blade row showing the lines of water


flow with a radial component [Fig. 3.4.6 (a)]. The outer portions of
the water droplets are deposited on the outer sidewall (shown at
point R). After impact, this water will flow along the sidewall until
it reaches the discharge point on the diaphragm, at which point it

298
Steam Path Damage Induced by Water

will either be detached and flow over into the rotating blade row, or
flow out over the vertical face of the diaphragm and along the cas-
ing. A similar mechanism is shown for a diaphragm with slanted
sidewalls [Fig. 3.4.6 (b)].

Figure 3.4.6Moisture deposition and flow to the outer side walls of a


fixed blade row.

Because of radial flow patterns, it is unlikely water will be


deposited on the inner sidewall to any considerable extent. How-
ever, it is possible that any water flowing under the stationary blade
row between the blade and rotor with leaking steam will migrate up
into the blade row. This is a particularly unpredictable mechanism,
and can only be determined to exist from flow patterns on the blades
when they are removed for examination at an outage.

In the case of rotating blades, as soon as moisture is deposited


from the steam flow onto the vane surface, it will experience high
centrifugal loading, causing it to flow outwards across the vane
faces. This effect is shown in Figure 3.4.7, where moisture is shown
to be deposited at positions R1...R4. From there, it flows outwards

299
Turbine Steam Path Maintenance and RepairVolume One

to the underside of the coverband band. At these locations, the mois-


ture experiences a combined force from centrifugal and steam flow
effects, causing it to have a substantially radial direction. At the
coverband band, the water will flow in an axial direction, coming
into contact with the coverband band under-face. This will continue
until, at point Zthe discharge pointit will be centrifuged out to
the casing or moisture-collecting device.

Figure 3.4.7Moisture deposition and


flow to the cover band in a rotating blade
row.

Basically, no moisture is deposited on the inner sidewall from the


steam flow, although it is possible some small amount could be pres-
ent there (but this will originate from the leakage under the stationary
blade row). As soon as this moisture is deposited on the blade lower
surface or inner sidewall, it will experience large centrifugal inertia
forces and flow outwards along the vane to the outer sidewall. In cer-
tain instances this water could detach from the blade root platform
(particularly if there are any flow discontinuities on surfaces, such as
the transition point from one blade platform to the next). In flowing
outwards, the steam flow forces would influence the moisture, and it

300
Steam Path Damage Induced by Water

is not possible to predict any flow directions. However, this quantity


would normally be small and of little interest in terms of predicting
possible damage scenarios.

Water flow along the casing


Because of the radial component in both the steam and the water
flowing in the steam path, centrifuged moisture continually adds a
film of water on the casing inner surface. This film flows from the high-
er to lower pressure areas of the casing. Due to the gravity effect, it
will also flow from the upper regions to the lower, from which points
it can be drained. Another influence on its flow is the windage effect
of the rotating blades. They set up forces that will attempt to force the
steam onto the sidewalls, above the rotating blades, to flow in the
same direction as the blades rotate. This is a small effect and consid-
ered to be of little consequence in considering damage mechanisms.

What is significant with this casing flow is when it migrates or


drains to the lower regions of the unit, it must be removed, as it has the
potential to cause significant erosive damage to the rotating blades.

MEASURING MOISTURE
DISTRIBUTION AND CONTENT
Until recently, the actual measurement of water content or dry-
ness in the steam path was particularly difficult to undertake, and reli-
able results hard to come by. However, recent technological improve-
ments have introduced methods that provide data allowing this to be
undertaken with acceptable levels of accuracy. It is anticipated that
as technology advances, these methodologies will improve, and there
will be even greater levels of confidence in the results.

301
Turbine Steam Path Maintenance and RepairVolume One

Wetness probes can be placed in the turbine to traverse the radi-


al direction and measure the moisture content at various locations.
Figure 3.5.1 shows a radial probe that was mounted on the down-
stream side of a row of exhaust stage blades before replacing the cas-
ing. This probe [Fig. 3.5.2(a) and (b)] had alternative positions at
entry to and exhaust from a large exhaust stage blade row.

Typically, a probe will use the light scattering technique


employing an argon-iron laser that projects a light into the flow path.
This laser light is reflected off the flowing droplets, and reflected into
a collecting device capable of counting/measuring their size. This
device then integrates the total to determine the moisture present. By
repeating this procedure at a number of radial locations, the total
moisture distribution can be determined. The results from such a test
are shown in Figure 3.3.4. STI/EPRI

Figure 3.5.1Laser probe at a last


stage blade discharge.

302
Steam Path Damage Induced by Water

Note:
CN11 - Generator North
CN12 - Generator North
CS11 - Generator South
CS12 - Generator South
CN0 - Generator North
CS0 - Generator South
STI/EPRI

Figure 3.5.2(a)The laser probe and its means entering


the low pressure section.

303
Turbine Steam Path Maintenance and RepairVolume One

STI/EPRI

Figure 3.5.2(b)The laser probe for moisture and droplet size measurement,
using light scattering and attenuation.

WATER REMOVAL
FROM THE STEAM PATH
All units in which steam operates with a moisture content have
the need for the control and removal from the steam path as large a
portion of this moisture as possible. It is also necessary for portions
of the units not operating entirely or exclusively in the moisture

304
Steam Path Damage Induced by Water

region to have provision for moisture removal. This is necessary so


that at a cold start, water present in the steam path can be drained
away before high rotational speeds are reached. The majority of
stages rely on the heating, evaporation, and carryover of the con-
densed moisture for its removal. This is a satisfactory method, but
operators must be sure that sufficient heat has been added so all
water has evaporated before high rotational speeds are reached.

There are three distinct methods of controlling the level of mois-


ture in a unit:

By controlling the steam conditions. This applies to both the


superheated and saturated steam cycles. Saturated cycles
include both nuclear and many geothermal

The removal of water from steam as it is transported from one


turbine section to the next. This is most effective in water-
cooled nuclear cycles where the steam entering the high-pres-
sure turbine already contains a small percentage of moisture
or has only a very small degree of superheat so moisture is
formed early in the expansion

In cycles using saturated steam, such as nuclear cycles from


water-cooled reactors and geothermal applications, the
steam is dried before entry to the unit by similar moisture
separating devices

The collection and drainage of the deposited moisture from


the individual stages

Internal collection and removal


For those stages that heating prior to start-up cannot effectively
dry, and those stages having moisture formed in and deposited on
them during operation, effective removal systems must be incorpo-
rated into the stage design.

305
Turbine Steam Path Maintenance and RepairVolume One

To achieve drainage, it is necessary to collect this moisture as


effectively as possible, pass it to a location where it can be drained,
and then provide an adequate passage for its removal. This drainage
must be completed without excessive steam blow down from the unit.

The majority of units employed on fossil-fuelled cycles have their


low-pressure sections designed to collect and drain moisture. For
units designed to operate on steam generated in the family of water
cooled reactors, moisture is present in even the highest-pressure
stages. Therefore, it is necessary to collect and drain from the high-
pressure section as well as from the lower pressure sections.

Moisture is collected at the casing inner surface by the incorpo-


ration of collection slots, or catchers, into the surface geometry.
There are three sources of water in these casing wall locations:

that which is deposited on the casing inner surface from the


radial flow effects of the expanding steam. This is a relative-
ly small quantity, but must be allowed for. In those stages
where there has not been sufficient elapsed time, after the
formation of water seeds, for these seeds to collect into large
drops, this may be the only source of moisture

water deposited on the stationary blade rowson both the


vanes and the outer sidewallsand flows with a large radial
component out to the casing. This is part of the total station-
ary blade row-deposited moisture that does not re-enter the
main steam flow as destructively large droplets

water centrifuged to the casing from the rotating blades. After


impacting with the rotating blades, a portion of this moisture
rebounds and will re-enter the main steam flow. However,
the major portion will be centrifuged out to the casing inner
surface

To collect as great a percentage of this moisture as possible, con-


siderable care is taken in designing moisture collection slots. They

306
Steam Path Damage Induced by Water

must allow the moisture to enter, be retained, and then drained to


the bottom half casing where it is removed. To achieve effective
drainage, the collection slots must be dimensioned to collect mois-
ture flowing along the outer sidewalls. The slot is optimally placed
above the moisture that is centrifuged from the rotating blade rows.
To achieve optimum axial placement, allowance must be made for
the differential expansion occurring in the steam path. Differential
expansion is load-sensitive, because the temperature of the various
portions of the steam path varies with the valve settings, which are
adjusted for the demand placed on the unit.

Wc

Wn

Figure 3.6.1The internal water 3.6.1


Figure catcher belt.
The internal water catcher belt.

As stated, the most common internal removal device, and that


employed by all manufacturers, is to collect a large portion of the
deposited moisture existing at the casing inner surface, and drain it
to a suitable point in the cycle by means of a circumferential water
catcher belt (often through a U tube). Figure 3.6.1 shows the basic
arrangement of a water catcher belt placed radially outwards from a
rotating blade. The method of constructing the collector belt can be

307
Turbine Steam Path Maintenance and RepairVolume One

seen. Here, at entrance to the catcher, an axial gap A is positioned


at a distance S (in the operating axial position) from the rotating
blade centerline to allow the moisture to enter. There is a relatively
large radial distance R above the entry point to prevent water
rebound, and then lips of depth D are produced at the belt entry
providing a drainage path for the collected water to drain to the bot-
tom dead center where it can be removed.

Note: The U-tube can, on load reduction, allow a portion of


the moisture to flash to steam, which will re-enter the steam path.
On load rejection it is possible a portion of the U-tube water will
re-enter the steam path as slugs possibly causing significant dam-
age (reference following section Water Ingestion into the Steam
Path).

This catcher (circumferential belt) removal system, shown in


Figure 3.6.1, is commonly employed in the latter stages of all low-
pressure sections, where large quantities of free moisture exist. These
designs are dimensioned to capture three separate sources of water:

Wc that enters the stage along the casing, and was not
removed in previous removal devices

Wn that is deposited on the stationary blade row, flows to


the casing and joins flow Wc

The catcher is also positioned so that Wr, centrifuged from


the rotating blade, is captured

These designs are also arranged to have a steam blow-down of


about 0.5% of the steam flow to help ensure effective removal of the
water.

While this drainage design is used in the turbine low-pressure sec-


tion, such a design can also be applied in the high-pressure expansion
in saturated steam nuclear units. In addition to the geometry shown in
Figure 3.6.1 an alternate arrangement, shown in Figure 3.6.2, can be

308
Steam Path Damage Induced by Water

used where the collector belt is located above the inlet edge of the
rotating blades. In this design the inlet nose of the rotating blade has
collection grooves, (shown in the detail) which capture the inflowing
moisture particles M, then provide channels for this moisture to be
centrifuged to the casing, where it will enter the catcher throat A.

Figure 3.6.2The water catcher belt, used with grooved inlet edge
blades.

309
Turbine Steam Path Maintenance and RepairVolume One

Moisture removal in the nuclear high-pressure sections is not


undertaken to prevent erosion on those stages, rather it has the pur-
pose of removing moisture from the expansion before it causes an
excessive efficiency loss. This internal separation is also used as a
preventative measure to reduce the duty imposed on the moisture
separator in the nuclear unit intermediate system. While moisture
quantities removed may be relatively small, they will in fact modify
the shape of the expansion line as shown in Figure 3.6.3, where the
enthalpy into a stage will increase by a small amount dH from
d1 to d2, and there will be a small increase in the stage entropy
of dS. Examining one stage it can be seen that while there is a shift
of the expansion line to the right, potentially lowering the level of
available energy, there will also be a small increase in the state line
efficiency because of the reduced moisture levels. Both of these
changes are small and would not normally be able to be shown on
the Mollier Diagram, but the effect is present.

Another method used to remove the moisture, is to provide a


drain in the outer wall behind each nozzle partition. This has been
shown to reduce erosion somewhat. However, the machining of the
individual drain and bleeder lines is relatively expensive. Another
method being used, but is too costly for many applications, is to
bleed the boundary layer on the stationary blade partitions and draw
away the moisture flowing there. This system has not been used
extensively since results have rarely been found to justify the manu-
facturing cost.

At locations in the steam path where steam is extracted for


regenerative feed heating, or other extraction purpose, if moisture
exists in droplet form, this steam extraction acts as a very effective
collection and removal device, removing a large portion of the mois-
ture present in droplet form.

310
Steam Path Damage Induced by Water

A Sat. X = 1.0
Xi
Pi

(a) dH

(b) d2
d1
(c)
dS

(d)

Pe
B

Figure 3.6.3The effects


Figure of internal
3.6.3
moisture
The effectsseparation
of internalin the highseparation
moisture pressure
section.
in the high pressure section.

External removal (saturated steam cycles)


Cycles employing steam at or near the saturated condition upon
entry to the turbine can, if the steam is to be removed during its total
expansion, employ separating devices external to the turbine. There
are various forms of separators all relying on the difference in densi-
ty between the steam and water particles to effect separation. The
methods of undertaking this separation are not a subject for this sec-
tion. However, the effect of these separators on steam conditions is
considered in greater detail in the following section, concerning sat-
urated steam cycles.

To remove this moisture, the steam must be removed from the


turbine after partial expansion, where it can be dried, and the water
drained. External separation is found suitable only for these cycles
where the steam when supplied to the unit contains transported
moisture, as in some geothermal applications and those units used

311
Turbine Steam Path Maintenance and RepairVolume One

with water-cooled reactors. In these cycles, steam is removed from


the high-pressure section, and passed through an intermediate sys-
tem, where it is dried before re-entry for the low-pressure expan-
sions. It is in this intermediate system that the moisture is removed
in a separator vessel.

Control of steam conditions


The moisture that exists at any point in the steam path is funda-
mental to the performance of the unit, and excess water quantities
can result in damage beyond the level anticipated by the designer.
The most convenient means of considering the influence of chang-
ing steam parameters on steam path water content is to review the
steam conditions, and changing conditions, as they are represented
on the Mollier Diagram.

The expansion lines for various cycles are shown in Figure 3.6.4.
These lines indicate the moisture levels anticipated by the designer,
and for which protective devices or water control systems are
designed into the steam power cycle. Changes in initial, reheat and
exhaust steam conditions will almost always affect operating mois-
ture levels and modify the potential for damage. Therefore, if a unit
has operated for a number of years with steam conditions within the
cycle that are off-design, through turbine, boiler, or condenser
inability to maintain design specified operating parameters (i.e., the
inlet, reheat, and exhaust steam conditions could not or were not
maintained) then the moisture levels will modify, and damage could
result. Under these conditions the units mechanical condition could
deteriorate, and should be examined for damage.

Superheated steam cycles. Initially superheated steam cycle con-


ditions for a typical fossil unit are shown in Figure 3.6.5. In this fig-
ure the design condition and variations are shown as heavy lines. The
initial design conditions are Pd and Td, point A. In the high-
pressure section the steam would then expand to pressure Ped at

312
Steam Path Damage Induced by Water

point B. Also shown in Figure 3.6.5 are variations from these initial
design inlet conditions from pressures Pu to Pl and temperatures
from Tu to Tl. It can be seen that any variations within these
ranges will not affect moisture, but will affect the conditions at B,
exhaust from the high-pressure section. At any flow quantity the inlet
to discharge pressure ratios will be substantially constant:
Pu/Peu = Pd/Ped = Pl/Pel

Tic
Pif T
Tif
Key: Pif
Pif
P = pressure
Fossil
T = temperature
reheat
X = moisture
Pcf
f = fossil
n = reheat
i = inlet
e = exhaust
r = nuclear reheat
o = nuclear non-reheat Entry to
d = design LP section
c = crossover pressure Pin

Tin Pcn

Sat. X = 1.0

Xi
TTD
Pin
Saturation
line

Nuclear
reheat
Xen

Nuclear
non-reheat
Xfr
Lines of
constant
moisture Px

Xnr

Xno

Figure 3.6.4Various Figure 3.6.4


cycles on the Mollier Diagram.
Various cycles on the Mollier Diagram.

313
Turbine Steam Path Maintenance and RepairVolume One

Pr
Tru
C
P Trd
Pd
u Tu Trl
A

Td
Tl Pr
Pl

Peu B
Ped

Pel
Sat. X = 1.0

Pci
Pcd
Pcu

Xi Pxi
Pxd
Xd
Xu

E
Pxu

Figure 3.6.5The superheated steam cycle.


Figure 3.6.5
The superheated steam cycle.

From the high-pressure section exhaust, in the reheat system to


inlet to the reheat section, there is a pressure drop in that system
(piping, boiler reheater section, valves, etc.) of Pr, and the steam
enters the reheat section at a pressure Pr, point C. Depending
upon the pressure drop Pr there can be some variation of reheater
inlet pressure. Therefore, the pressure at inlet to the reheat section is
dependent upon the initial pressure to the turbine at A and the
pressure drop in the reheat components comprising the cycle.
Therefore, the re-energized steam enters the reheat turbine section at
C with a pressure Pr.

314
Steam Path Damage Induced by Water

However, despite these possible differences from HP turbine


exhaust to reheat section inlet, the major parameter influencing low-
pressure exhaust conditions, and therefore steam path moisture con-
tent is the reheat temperature Tr at C. The effect of varying this
temperature from Tru to Trl can be seen in its effect on the pres-
sure at which the expansion line crosses the saturation line D, the
steam path and final moisture content at E, when the condenser
pressure is Pxd. The variation of exhaust dryness content varies
from the design-anticipated value of Xd by amounts Xu to Xl.

Saturated steam cycles. In saturated steam cycles such as the


nuclear and many of the geothermal applications, steam enters the
turbine with initial moisture content (shown as condition A in Fig.
3.6.6, which represents the expansion lines for such saturated con-
ditions). Steam produced by the water-cooled nuclear reactors has
conditions of about 1,000 psia with small initial moisture content of
about 0.25%, Xdi. The moisture content before entry to the turbine
having been reduced by the removal of any free moisture it might
have been transporting from the nuclear boilers.

After admission to the high-pressure section of the turbine, the


steam expands to the design selected pressure Pi, at B. This steam
then exhausts from the turbine, and passes through an intermediate
system to improve the steam conditions by drying it to condition
C, and then either reheating it to condition D, or returning it to
the low-pressure sections where it expands in the non-reheat cycle
to condition E. For the reheat application the steam is reheated
using live steam extracted from the nuclear boiler (and possibly uti-
lizing partially expanded steam removed from the high-pressure tur-
bine). Reheating improves the steam condition to D. Steam is then
allowed to expand in the low-pressure sections to condition F.

At entry to the high-pressure section of the turbine, the steam has


a condition established by the performance of the nuclear boiler,
and more importantly in terms of moisture by the performance of the
moisture separator, which if not working correctly will lower the

315
Turbine Steam Path Maintenance and RepairVolume One

inlet conditions to pressure Pd, moisture content Xdl. The steam


then expands, and will achieve the same pressure ratio across the
high-pressure section. This will decrease the dryness content in the
high-pressure section exhaust steam from Xe to Xel.

Pi
Pi TTD
Td
Tid

Tidl
A Xdi
Xdl Xi
Xil C
Sat. X = 1.0
Pd
Pid

B
Xe
Pi

Pxd

Xdn
F
Xdnl Xdr
E Xdrl

Figure 3.6.6The nuclear


Figurecycle on the Mollier
3.6.6
Diagram.
The nuclear cycle on the Mollier Diagram.

In passing through the intermediate system there will be a pres-


sure drop Pi. Also the performance of the external moisture sep-
arator will, if operating at design specification, improve the condi-
tion to C, dryness Xi. If however, the separator is not operating

316
Steam Path Damage Induced by Water

correctly then some greater moisture content at dryness Xil will be


achieved. The effect of this on the non-reheat cycle can decrease the
exhaust dryness from Xdn to Xdnl.

If live steam reheating is to be used in passing into the reheater


portion of the cycle, energy will be used undertaking the additional
heating to evaporate moisture, improving the steam condition from
Xil to saturated, rather than from Xi to the saturated condition. A
degree of superheat will then be added by the reheating steam
improving the steam temperature to Tid, which is less than the sat-
urated steam temperature corresponding to inlet pressure by an
amount termed the terminal temperature difference (TTD). The
maintenance of the TTD will help maintain the exhaust dryness at
Xdr. An increase in the TTD will decrease the dryness from Xdr
to Xdrl.

In the simple geothermal application, shown in Figure 3.6.7,


steam will enter the unit with a condition A, pressure Pi, and dry-
ness Xi. (The inlet pressure on the majority of geothermal units is
in the range 100psia+/-.) This steam will expand with normal effi-
ciency to condition G, pressure Px, and at this condition, the
dryness will be Xgd. In the geothermal cycle, the major influenc-
ing factor is the effectiveness of the moisture separator ahead of the
inlet valvesif not operating with design effectiveness, could give a
dryness to the turbine of Xiu at entry, causing an increase in the
final moisture content from Xgd to Xgu.

Exhaust steam conditions. As the condenser reduces exhaust


pressure, there will be an increase in the available energy, and the
last stage discharge moisture content will increase with decreasing
pressure. The extent to which damage occursas a consequence of
this additional moisture that forms in the last stage due to the reduc-
tion of exhaust pressureis dependent upon the location of the
moisture and the particular form of damage.

317
Turbine Steam Path Maintenance and RepairVolume One

Xi
A
Sat. X = 1.0

Xiu

Pi

Px

Xgu G
Xgd

Figure 3.6.7The
Figure geothermal
3.6.7
cycle. The geothermal cycle.

Damage, and its dependence upon moisture content, will be


considered in greater detail in sections following.

It is clear from this discussion of steam conditions, these will have


a considerable effect on the moisture content in the various portions
of the unit. It is also clear there are pieces of equipment contained
within the cycles that will also have a significant impact on the final
moisture content, and the moisture throughout the steam path.

The effects of internal moisture removal


on steam path conditions
The means of moisture removal by collector belts was discussed
earlier, and this action of collecting and removing was shown on the
Mollier Diagram Figure 3.6.3 for a high-pressure section, where the
removal was undertaken from efficiency considerations. In the low-
pressure section the quantity of moisture removed is more substantial,

318
Steam Path Damage Induced by Water

and the effect on steam path conditions more pronounced. The effect
of such removal is shown in Figure 3.6.8 for a low-pressure section.

The effect of internal moisture separation is to reduce the mois-


ture content of the working fluid at the stage position. This is indi-
cated on the Mollier Diagram by expansion line discontinuities. On
the expansion line shown in Figure 3.6.8 the steam enters the low-
pressure section with a condition Pm/Xm, for a fossil or nuclear
unit. In this low-pressure section there are three stage positions
where moisture is removedn1, n2, and n3. At each stage posi-
tion the pressure remains unchanged, but the level of moisture is
reduced. The net effect of this three-stage separation is to reduce the
exhaust, or final dryness from Xf to Xe.

Reheat
section
expansion
Xm
Sat. X = 1.0

Pm

b
a
n1 d
c
n2 f
points
Stage

e
Px
n3

Xe

Xf

Figure 3.6.8The
Figureeffects
3.6.8 of inter-
nalThe
moisture
effectsseparation
of internal on steam
moisture
conditions
separationin on
a low pressure
steam sec-
conditions
tion. in a low pressure section.

319
Turbine Steam Path Maintenance and RepairVolume One

In the first removal stage the condition is increased from a to


b, drying the steam. A similar effect can be seen for the other two
stages. These stages will have a total effect of modifying the final dry-
ness from Xf to Xe, and having the same effect on steam condi-
tions as discussed for Figure 3.6.3.

Sat X = 1.0
A

Pi
Inefficient
Expansion
Line

Design
Expansion
Line
Xd D
Xb

Po B dS d

Figure 3.6.9The
Figureeffect
3.6.9 of inefficien-
cy onThe
moisture
effect content.
of inefficiency on
moisture content.

The effects of steam path efficiency


on moisture content
From its initial start to the following outage, steam path efficien-
cy will deteriorate. This deterioration will cause an increase in the
expansion line entropy, making less energy available for conversion
to work. However, as shown in Figure 3.6.9, for an expansion from
pressure Pi to Po this deterioration will also increase the final
dryness form Xb to Xe. This drying is a reflection that the addi-
tional heat generated by frictional heating (inefficiency), will be
returned to the working fluid, and will in effect evaporate some of

320
Steam Path Damage Induced by Water

the water contained in the steam. Similarly any increase in steam


leakage quantities (which reduces efficiency), will mean the energy
not removed from the leaking quantity will mix with and raise the
total energy level of the working fluid. Both effect drying the expand-
ing steam.

Similarly if any stage suffers mechanical damage, its efficiency


will degrade. This effect is shown in Figure 3.6.10, where steam
expanding from pressure Pa to Pb in a damaged stage will
increase the stage discharge entropy by an amount ds, and lose
enthalpy for conversion to work by an amount h. In doing this the
expansion line will shift from a-B to a-b-D, and the end effect on
moisture content will be to dry the steam from Xb to Xd. The
effect on final moisture content can be seen on the Mollier Diagram.
The result of any frictional losses will be a small increase in the stage
reheat.

Pi Sat X = 1.0
A

Design Damaged
Expansion s stage
Line
a h
pa b
c
pb

Po

D Xd
B Xb

Figure 3.6.10The
Figureeffect of stage
3.6.10
mechanical
The effect ofdamage on moisture
stage mechanical damage
content. on moisture content.

321
Turbine Steam Path Maintenance and RepairVolume One

The loss in efficiency of the entire expansion shown in Figure


3.6.9, or a single stage shown in Figure 3.6.10 can be found from:

dH . m
kW =
3412.14
h . m
kW =
3412.14

where:

kW = The power loss in kilowatts


m = The steam flow in #/hour

MOISTURE-INDUCED DAMAGE
Moisture in the steam path is capable of causing extensive
mechanical damage to the various components that comprise it. This
damage can both degrade efficiency and also force the unit from
service for extensive periods to repair and/or replace parts. There are
four major forms of damage to be considered, caused either by water
formed in the steam path, or intentionally sprayed into it for cooling
purposes:

Moisture-impact erosion. This is a form of erosion occurring on


the inlet edge suction surface of blades, in their outer regions. This
damage occurs as a consequence of the high impact force develop-
ing between the moisture particles and the blade material.

Trailing-edge erosion. This is a form of erosion that occurs on the


discharge portion of last stage blades near the root section. Damage
is a consequence of re-circulation of steam and water carried into
the blades from the exhaust casing.

322
Steam Path Damage Induced by Water

Wire-drawing erosion. Damage occurs when moisture, on one


side of a metallic joint is forced by a high-pressure differential to
flow between the joint faces, removing material by a washing/cavi-
tation process.

Washing erosion. Washing erosion occurs as material loss from


high velocity, possibly high-pressure steam flowing across a surface
and removing material.

The following sections will consider these types of damage in


greater detail, and where possible explain the material loss mecha-
nism or process.

MOISTURE-IMPACT EROSION
This is the most common form of water damage in the turbine
steam path. In order for this erosion to occur, the moisture must
accumulate into considerably larger drops than exist from the for-
mation of water seeds by nucleation. The damage results from the
impact forces developed between moisture droplets carried by the
steam and the blade.

During its passage between the blades, a portion of this trans-


ported moisture will be deposited upon the various elements of the
steam path, and flow across their surfaces. This water flows as a film
across the surfaces into regions where the film collects into larger
puddles. They are eventually torn from the surfaces and re-enter the
main steam flow, moving at considerably lower velocities.

This re-entrained moisture torn from the stationary portion (nor-


mally the stationary blade elements), causes impact erosion upon
collision with the leading edge of the rotating blades. Typical of the
material loss is that shown in Figure 3.8.1, where craters on the

323
Turbine Steam Path Maintenance and RepairVolume One

leading edge, and the extent of surface penetration back along the
suction face is shown. Figure 3.8.2 shows a micro-section through
a moisture impact eroded surface, defining the type of craters formed.

Figure 3.8.1Moisture impact erosion at a blade outer


sections.

324
Steam Path Damage Induced by Water

Figure 3.8.2Micrograph of a moisture impact eroded surface.

The mechanism of impact erosion


and material loss
The droplets formed in the steam path are not of themselves large
enough to cause erosive damage upon impact with the metallic sur-
faces of the rotating blade elements. These small moisture particles
must therefore aggregate to droplets of a much larger size before the
impact forces developed between them and the rotating blades are
sufficient to cause surface material rupture and loss. The previous
section Moisture Deposition and Figure 3.4.5 shows the mechanism
of water agglomeration on the discharge tail suction face of the sta-
tionary blades.

The water film flowing down the pressure face of the stationary
blade is drawn around the discharge tail into the lower pressure
region of the row. This is due to the separation of the boundary layer
on the suction face of the profile, at or near the passage throat, and
the lower pressure this induces in this region. At this point the film

325
Turbine Steam Path Maintenance and RepairVolume One

meets, and mixes with a small quantity of water flowing across the
suction face. On the discharge tail suction face, the water collects
into large droplets that grow until they are eventually torn from this
location, and are carried into the following row of rotating blades
(Fig. 3.4.5).

It is these large droplets, torn from the stationary blade trailing


edge that cause the impact erosion damage on the inlet edges of
the following row of rotating blade elements. Due to their greater
mass, these droplets once torn by the steam forces from the suction
face, accelerate from zero velocity, but cannot approach the veloci-
ty of the steam, which is high in the discharge region. Therefore, their
impact with the rotating blade inlet edge is at an oblique angle,
shown in the velocity triangle Figure 3.8.3. Upon impact large pres-
sure forces are developed between the blade surface and water
droplets.

Figure 3.8.3 shows the steam velocity triangle for the tip section
of a large rotating blade, having a high tangential velocity. These
steam velocity vectors have a suffix s. This velocity triangle is typ-
ical of the tip region of a large rotating blade row. The blade has a
high tangential velocity U, and the steam discharges from the sta-
tionary blade row at a velocity C1s, from a stationary blade with a
discharge angle 1. These velocities give a relative inlet velocity at
inlet to the rotating blade row of W1s, at a relative angle 1s.
Superimposed on this steam vector diagram are the velocity compo-
nents of the water droplets, suffix w. The water droplets, starting
from rest on the suction surface of the vane, have a relatively small
velocity on discharge from the stationary blade row, being acceler-
ated to a velocity C1w by the flowing steam. These droplets dis-
charge from the stationary blade row at the same discharge angle
1. The blade tangential velocity is unchanged at U. The vector
diagram for the water droplet is then shown having a relative inlet
velocity to the rotating blade row of W1w. The water droplets enter
the rotating blade row with a relative Shadow Angle .

326
Steam Path Damage Induced by Water

W1w 1
C1w
W1s 1s C1s

Figure 3.8.3
Figure 3.8.3The velocity triangles for steam s and water w from a wet stage
The velocity triangles for steam "s" and "water w" from a wet stage fixed blade row.
fixed blade row.

These large droplets torn from the stationary blade row discharge
edge, are normally too large (about 4x10-3) to exist in the steam
environment present at that location. Therefore, the drops break
down into droplets to a size they are able to sustain. These droplets
are still considerably larger than the seed elements (about 4.6x10-8
inches) formed by nucleation.

Figure 3.8.4Water drop size distribution (a), and their velocity


triangles (b).

Shown as Figure 3.8.4(a) is a diagrammatic representation of the


variation of droplet sizes as they enter the rotating blade row. These
droplets, having a spectrum of sizes, are accelerated to different

327
Turbine Steam Path Maintenance and RepairVolume One

velocities C1w by the transporting steam shown in Figure 3.8.4(b).


Such droplets achieve different relative velocities W1w and shad-
ow angles depending upon their size. They impact with the
rotating blades, but penetrate the blade cascade in an axial direction
to different levels. In Figure 3.8.4(b) the velocity diagrams for three
different drop sizes, L, M and S, can be seen. Their absolute
velocities are C1w, achieved after detaching from the stationary
vane and being accelerated by the steam flow, which is in the same
direction. Their shadow angles are , and their penetration in the
axial direction is shown in Figure 3.8.5. As a consequence of their
relative velocity W1w, their penetration upon entry to the rotating
blade row is dependent upon size, which influences the velocity
C1w, and the impact point on the rotating blade inlet edge.

Inlet
edge
W1w

Figure 3.8.5Showing the shadow angle


Figure and the degree of penetra-
3.8.5
tion into the rotating
Showingblade row. angle ' ' and the degree of
the shadow
penetration into the rotating blade row.

Upon impact between the water droplet and rotating blade inlet
edge, there are two factors that can account for, or contribute to the
erosive damage occurring. First, there is the initially high intensity
pressure developed between the droplet and blade material at the

328
Steam Path Damage Induced by Water

impact point. Second, there are the high radial flow velocities of the
water outwards from the point of impact.

Consider these two effects and how they would contribute to sur-
face material loss. If the impact of the high velocity water particle
and an elastic/plastic brittle material is considered, the material
deforms to an elastic crater under the impact forces [Fig. 3.8.6 (a)].
The magnitude of the stresses induced in the material are sufficient
to produce hairline cracks that form at and run around the shoulder
of the indentation. After impact, the material recovers substantially
retaining its original position. These hairline cracks will now have
closed, but there is a circular fracture zone existing around the
impact point. Shown in Figure 3.8.6(b) is a macro photograph of an
aluminum target after impact by a high velocity mercury drop fired
at the target. The circular crack zone can be seen.

Stress Produced by
p = Pressure stress.
t = Tensile stress.
s = Shear stress.

p
p p
s s

s s

Figure 3.8.6(a)The stress produced


Figure in (a)
3.8.6 a surface as a con-
sequenceTheof stress
an impact.
produced in a suface as a consequence
of an impact.

As a consequence of this initial impact there is probably no, or


minimal, material loss from the surface. However, subsequent
impacts in the region of the initial impact will cause water to be

329
Turbine Steam Path Maintenance and RepairVolume One

present on the surface, again causing impact indentations.


Immediately after these subsequent impacts, the water flows radial-
ly outwards. There are three aspects of this flow that can cause or
contribute towards the erosive material loss. These are the high
impact forces between the radially outflowing water and surface
irregularitiesmany such irregularities having been formed at the
hairline crack points. The high shear forces developed between the
flowing water and surface material, and finally any cavitation effects
of the high velocity water flowing across the profile face.

1 m.m.

Figure 3.8.6(b)The fracture rings" after impact of a mercury pellet with a hot
pressed aluminum target, at a velocity of 1400 ft/sec.

330
Steam Path Damage Induced by Water

Methods of protection
The methods of protection currently used against moisture
impact damage can be considered to employ two main principles
that can be applied simultaneously:

The collection and drainage of as much moisture as possible,


which is present in, and deposited upon the turbine elements

The deposition, production, or attachment upon the areas of


the blade vane surface likely to be affected (a layer of hard
resistant material that is better able to withstand the impact
forces developed by the water droplets)

A third method is to increase the axial gap between the station-


ary blade discharge and the rotating blades inlet edge. This increased
gap permits a greater distance, and therefore more time for the mois-
ture droplets to be accelerated to the velocity of the steam, reducing
their velocity difference relative to that of the blades. This method is
not employed extensively, as the cost of increasing the axial length
(compared to the losses in efficiency and the increased shaft length)
more than offset any potential gain from reduced erosion.

Moisture removal. Moisture removal by internal (water catchers,


etc.) and external means in the intermediate portion of the cycle
were considered in the previous sections Internal Collection and
Removal and External Removal.

Erosion resistant inlet edge. The most extensively used method


of erosion protection, is producing on the inlet edge of the vane in
the regions subject to erosion a hard resistant surface. There are sev-
eral methods of providing and then attaching such a resistant sur-
face. These include the following:

Braze attached resistant material. The braze attached shield is


the most common method of providing protection. The shield
can be of many forms. Figure 3.8.7 shows four possible

331
Turbine Steam Path Maintenance and RepairVolume One

designs. The form of the shield is selected to provide protec-


tion, and more recent designs, to preserve the aerodynamic
form of the vane at the point of attachment. Such contouring
allows the shield to attach within a preformed recess and pre-
serve the aerodynamic form of the profile. In Figures 3.8.7(a)
and (b) the shields are of relatively simple design, and make
no attempt to maintain the aerodynamic form of the vane.
There will then be some degree of boundary layer separation,
and efficiency loss with these forms. Figure 3.8.7(c) and (d)
employ more complex shield forms, fitted into recesses, or
specially designed access slots at vane the inlet. These shields
maintain the nose profile.

Figure 3.8.7Various nose forms of hard metal erosion shield.

Difficulty can be experienced with the braze attached shield,


if the procedures for attachment are not followed in detail.
The melting point of many suitable filler materials is close to
the austenitising temperature of the blade material. Failure to
control temperature can cause embrittlement of the vane at
any point of overheating. This can ultimately lead to prema-
ture failure of the blade material.

For brazed attachment, the shield can detach in operation if


the brazing joint is not adequate. Shield detachment leaves

332
Steam Path Damage Induced by Water

the blade tip material exposed, and therefore subject to a


heavy rate of material loss.

Another difficulty encountered with the brazed shield has


been that the softer braze material has been subject to heavy
moisture impact erosion, and some undercutting from the
moisture. It is possible however to design the shield so it is
partially self-protecting at the braze point.

Weld attached resistant material. Another method of attach-


ing shields is by welding. Control of the weld attachment
process is critical. Cracks have developed in the heat affect-
ed zone (HAZ) of the shield/blade interface, and while the
majority of these cracks have run forward towards the inlet
edge, some have regressed into the blade parent material. In
the event a crack does develop some form of corrective
action is required. This situation cannot be left uncorrected.

Thermal Hardening. Thermal hardening is undertaken using


either direct flame impingement or induction heating of the
blade material. With such a hardening system, process con-
trol is critical, and stringent observance of procedures is
mandatory. If this process is not controlled normally cracks
can initiate in the HAZ, causing the outer portion of the
blade to detach.

Figure 3.8.8 shows the section of a thermally hardened inlet


nose displaying the effects and results of a hardness traverse.
The base material had a design specified Brinell Hardness
(BHN) of 260-275, and the material as seen from the base
material readings fell into an acceptable range of values. At
the transition HAZ, there was a dramatic change of hardness
to the 400-415 BHN range (one reading taken too close to
the profile edge gave a questionable reading). The hardened
region ends abruptly at the HAZ.

333
Turbine Steam Path Maintenance and RepairVolume One

Laser hardening. A further development in the hardening of


the blade vane material is by the use of lasers. This is not as
common as other methods, but does promise to be placed
into greater use as results justify its application.

Figure 3.8.8The effects of thermal hardening of a vane inlet edge. In (a) is shown a
section through the hardened nose, and in (b) the results of a hardness traverses.

334
Steam Path Damage Induced by Water

Secondary (concentrated) erosion


If a blade that has been in service in the moisture region for an
extended period is examined, there are often regions where the local
depth of penetration by erosive action is far greater than in the sur-
rounding material. This localized heavier damage is caused by the
occurrence of obstructions or collection points at different up
stream locations within the steam path. At these points moisture can
collect or concentrate until it is suddenly torn off by the main steam
flow, or thrown off by centrifugal action, causing heavy localized
erosion on the following rotating blades. Figure 3.8.9 shows this
concentrated or secondary erosion.

Figure 3.8.9Concentrated or secondary


erosion.

335
Turbine Steam Path Maintenance and RepairVolume One

Careful attention to detail during the design phase can help rec-
ognize the potential for this form of damage, and help eliminate or
minimize the potential for it to occur. Prototype units should always
be examined early in their operating life, to determine if the early
stages of such damage are present. If damage has occurred or there
are indications of such concentrated penetration, the design should
be reviewed, and remedial action taken before the damage has pro-
gressed to the point its effects are significant, and before components
must be replaced prematurely.

In a well-designed stage, there should not be a need to replace


blades during a normal operating life of 200,000 hours due to mois-
ture impact erosion. However, concentrated erosion can force a unit
from service for an extended outage, with the need to replace the
affected row.

Off-shield erosion
During the design phase, the dimensional requirements of the
erosion shield are predicted from an analysis of the stage operating
conditions, and those anticipated under worst conditions, includ-
ing prediction of the shadow angles likely to be encountered. Using
this data, a prediction is made as to the extent of erosion, the degree
to which it is expected to occur on the blade, and therefore the level
of coverage required by a protective shield.

Low-pressure blading is designed for a range of applications, a


variety of stage inlet conditions, and a range of blade loadings
(pounds of steam/water flow in the stage). The low-pressure exhaust
blades can be subject to a range of moisture impingement probabil-
ities. Last stage shield requirements are designed to accommodate
this range of steam and water entering the rotating blade, and are
normally designed to be adequate under the most severe conditions
the blade is expected to encounter. However, it is a common oper-
ating experience that occasionally unique features of operation, or

336
Steam Path Damage Induced by Water

characteristics of a particular unit, can cause unusual patterns of


material loss. These unique situations can cause the shield to be
inadequate and erosion can be noted to occur to the extent the life
of the erosion shield and/or blade can be placed in jeopardy.

Whenever a unit is opened for inspection and maintenance,


blades should be examined for the depth of erosion, and also for any
unusual off-shield erosion. If it is considered the level of damage
is progressing to the extent excessive damage can be anticipated and
the reliability of the unit could be affected, the condition should be
monitored either by measurements, making casts of the damage, or
photographs. This monitoring should be done in such a manner the
rate of penetration or extent of damage can be predicted and judg-
ment made, even though subjective, of when corrective action could
be required. Remedial action must be planned for and taken before
the blade is destroyed. Destruction of the blade is not of itself cata-
strophic, however, consequential damage can be very expensive to
correct.

The various forms of off shield erosion that a blade should be


examined for consist of the following:

Beyond shield erosion. In this case the impact erosion extends


back, beyond or behind the shield coverage into the base material.
Such damage is shown in Figure 3.8.10. Often, this is not significant
as the erosion may only be of a very light nature and therefore will
not endanger the integrity of the blade. Should this occur however,
it is advisable to note its extent and radial location and then monitor
further deterioration of the vane. Should this type of damage be
found early in the operating life of the unitbefore 10,000 operat-
ing hoursand occur at one or two radial locations, it would be
advisable to monitor the condition of the shield in this region, since
this could be symptomatic of heavy localized, or secondary erosion,
which will occur there.

337
Turbine Steam Path Maintenance and RepairVolume One

Figure 3.8.10Erosion beyond the erosion shield.

It is also possible with a shield attached by brazing, that early


erosion behind the shield is simply removing the excess soft braze
material. This type of braze erosion should not be confused with ero-
sion of the blade material.

Below shield erosion. Normally, erosion does not occur below a


blade threshold tangential velocity of about 800-850 feet/second (50"
diameter at 3600 rpm). The shield will normally be designed to
extend down the blade inlet edge until a point is reached on the pro-
file at which this velocity does not occur. There are occasions when
erosion is found below a shield and this progresses into the blade
material to the extent the shield can be undercut and severe damage
occur due to the stress concentration effect this material loss causes.
Material loss in such a location can also modify the natural frequen-

338
Steam Path Damage Induced by Water

cies of the blades and move their frequency into a dangerous range.
In such a situation shown in Figure 3.8.11, it is necessary to consid-
er the extent to which the shield can be produced or modified to
cover the lower radial portions. It would also be advisable to consid-
er if by modifying moisture flow patterns some corrective actions on
the preceding stationary blade row would help alleviate the problem.

STI/EPRI

Figure 3.8.11Erosion below the lower end of the


erosion shield.

Erosion between shield segments. Many units, particularly those


of an older design, employ an erosion shield produced from two or
more individual segments. At the point where the shield segments
butt, it is possible for erosion to occur and proceed into the softer
braze or blade material. The shield is generally arranged in its radial
direction so overlap of shield pieces occurs and the radial flow

339
Next Page

Turbine Steam Path Maintenance and RepairVolume One

direction of the moisture particles tends to cause less severe damage.


However, such damage does occur. Figure 3.8.12 shows the extreme
of this type of erosion, where the outer segment of a shield has
detached, and the resulting depth of penetration in the outer blade
portion that was consequently exposed is being destroyed. In many
instances, significant material loss can occur on the two ends of the
shield, without its detachment.

If damage of the type shown in Figure 3.8.12 occurs, this can, if


it is only on one or two elements, affect the dynamic balance of the
unit, as the mass is being removed from the outer diameters of the
vane where its centrifugal mass kinetic effect on balance, is more
significant.

Turbine Blading (UK) Ltd

Figure 3.8.12Heavy erosion into the blade base material after loss of the erosion
shield. The shield having detached at the radial position where two segments butt.

340
Chapter

4
Operational Events
Giving Rise to
Steam Path Damage

INTRODUCTION
The steam path can be subjected to a number of operational
events that degrade the overall quality of the unit, and so degrade
performance, both in terms of efficiency and availability. Under cer-
tain circumstances these damaging mechanisms can be sufficient to
force the unit from service, and degrading influences can result in
considerable expense to the owner.

Not every machine is subject to each of the damage mecha-


nisms that follow. But each machine can be if the operational con-
ditions are such that the opportunity for their occurrence exists.

This chapter will examine the most common of these, and where
possible provide an explanation of their characteristics, some brief

397
Turbine Steam Path Maintenance and RepairVolume One

explanation of the circumstances that allows them to occur and pro-


vide some guidance to their avoidance, if such avoidance is possible.

Some of these mechanisms are well understood and some


allowance is made for them in the original configuration of both the
cycle and the details of the turbine unit. Other mechanisms are often
unanticipated, and can be particularly present in units of a prototype
design when the boundaries of design are extended and experience
extrapolated to provide guidance.

FOREIGN OBJECT
IMPACT DAMAGE
There are many instances when the blade vanes, both stationary
and rotating, are damaged by impact with solid objects that have been
generated within, or gained access to the steam path. There are a
number of sources for these objects, which can be of various sizes
depending upon their origin. The damage they cause will depend
upon their mass, their location in the steam path, and the environ-
mental temperature that give the material different mechanical prop-
erties.

These objects have the potential to effect the operation of the


unit, and must be given consideration. They include:

The larger objects, if generated within the steam path, may


be the mass of an entire blade, which having suffered failure
in the root section, detaches and rebounds within the
remaining blades

This can cause significant consequential damage, destroying a


number of the blade elements within the row, some of which may
also detach. Such failures, when large blades are involved, can often

398
Steam Path Damage Induced by Operational Events

affect the dynamic balance of the unit, introduce dangerous level of


rotor vibration, and force it from service.

Often when smaller blade elements fail, they do not affect bal-
ance, and there is no indication such a failure has occurred. The
transient on any vibration-recording device may be so small it is not
noticed, and the unit will continue to operate in the same manner
with no evidence of failure. That a failure has occurred is often not
detected until the unit is opened at the next inspection/maintenance
outage.

The objectswhether or not they affect balance and require


shutdown to take corrective actionwill almost certainly
cause mechanical damage if theyre free in the steam path

The exceptions to the occurrence of this impact damage are the


possible failure of a piece of stage hardware, such as a coverband or
tie wire. When they become loose, they can wedge at the casing
inner surface, possibly at a diaphragm or steam or water extraction
slot, and not re-enter the steam path. This is rare, but can occur.
Another possible exception to the occurrence of damage is when a
last-stage blade (or a piece of last stage hardware in any section) fails
and the generated missile is carried directly into the exhaust. For a
low-pressure section, the missile can be carried directly into the con-
denser tube.

When a piece of debris re-enters the steam path and impacts


with the blades, it causes mechanical damage that introduces an
impact pit and/or causes material rupture. The extent and form of
the damage depends upon the size of the piece, the temperature of
the surface with which it impacts, and the point of the blade or other
surface on which it impacts.

Should the impact occur on a thin section of the profile, the


deformation that results is likely to be more severe and damaging to
the aerodynamic form, than if a more robust part is struck.

399
Turbine Steam Path Maintenance and RepairVolume One

When objects too large to pass through the blade passages are
generated, they may wedge between the stationary blade rows.
Often the rotating blades cut the large piece into a number of small-
er segments, which rebound in the axial gaps between the stages
until they are reduced to a size small enough to pass downstream.
This chopping action does considerable damage to the blade inlet or
discharge edges (whichever is slicing the large particle), and nor-
mally results in the replacement of the entire blade row. The pieces
passing downstream then impact on the following blade rows, caus-
ing damage to them.

Surface roughening and efficiency loss are the consequences


of the objects impact with the vane surface

This is not, however, the only form of loss caused by object


impacts. There are other types just as damaging to unit performance.
The major factors introducing losses are:

Surface roughening associated with the production of impact


craters on the vane surface

These craters appear at the point of impact, which for small


objects tends to be at the inlet on the suction face. However, as the
objects become larger, there is no method of generalizing location
or severity. The craters tend to make the boundary layer separate
from the vane surface prematurely, and in so doing create vortices
that will then exist throughout the expansion. Where the boundary
layer separation occurs, will depend upon the axial location of the
crater.

By impacting with the inlet nosewhich can be relatively


thin on many designs, particularly those of an older vin-
tagethe nose is deformed

This modifies the inlet angle causing an incidence loss. The


impacts can also be sufficiently damaging that they close down the
inlet, making it more difficult for the steam to enter the blade passage.

400
Steam Path Damage Induced by Operational Events

Impacts on the discharge edge from objects existing in the


axial gap between the stationary blade discharge and rotat-
ing blade inlet will often close down the stage discharge area

This will modify pressure from the stage and modify the stage
enthalpy and blade velocity ratio. It will also modify the discharge
angle from the optimum. Damage to the discharge edge can modify
the discharge area, the steam flow, and the pressure distribution
throughout the steam path.

Debris near to the blade tips can cause damage to any seal
or leakage control devices and deform them, increasing leak-
age that occurs there

Free hard objects in the steam path, whatever their origin, are
destructive and have the potential to degrade efficiency, and possi-
bly cause the unit to be forced from service immediately. They may
also introduce a condition that will deteriorate, possibly causing fail-
ure at a later date, or extending an outage while replacement parts
are obtained.

SOURCES OF THE
IMPACTING OBJECTS
To cause damage within the turbine steam path, missiles must
gain access, or be present in some manner. The most likely causes
for them gaining access, or being present follow:

The result of mechanical failure of some portion of the steam


path. For example, when a component or component part detaches
and becomes loose within the blade system, the resulting debris will
be centrifuged to the casing and rebound, most likely re-entering the
steam path and possibly making many damaging impacts before its

401
Turbine Steam Path Maintenance and RepairVolume One

dangerous effects cease. Such debris can find several final locations
in the unit:

Lodging in the casing in a feed heating extraction line or sim-


ilar location

Lodging in a diaphragm or stationary blade row

Being chopped into smaller pieces and distributed through-


out the steam path

Being carried into the low-pressure section and transported


into the condenser. In many instances, larger pieces carried
into the condenser will cut or severely damage tubing. This
will possibly cause other problems associated with the
ingress of cooling water, which will be chemically unsuitable
for use in the turbine system

Being carried over into a feed heater, steam extraction pip-


ing, feed heater, or other extraction line

Debris carried into the unit from the boiler. While this material
generally consists of weld bead (or even stick ends resulting from ini-
tial construction and post-repair problems), other pieces have been
found in steam lines. The damaging effects can be minimized by the
use of fine mesh screens for short periods, after initial start up and
after return-to-service, when repairs have been made. There is an
additional pressure drop (and efficiency degradation) associated
with the use of the fine mesh screen, but the removal of weld bead
is important. The fine mesh screen need only be used for about six
to eight weeks.

Debris entering the unit from some external source. This is


unlikely, but there are instances in which components external to the
steam path proper can fail and enter the blade system.

402
Steam Path Damage Induced by Operational Events

Parts left in the unit during an outage. Good maintenance prac-


tice makes this uncommon, but often small pieces such as nuts,
bolts, and small tools will be dropped into the unit during an outage
when the unit is being inspected and possibly refurbished. These
pieces are most often metallic, and have the potential to cause sig-
nificant damage upon restart.

Of the various sources of steam path debris, the generation of a


missile from steam path components themselves is the most com-
mon and most likely to cause the most severe structural damage in
terms of requiring component replacement. Such missiles are the
result of damage in their stage, and may cause consequential dam-
age that may be greater than the damage suffered by the stage in
which the initial failure occurred.

Formed internally
This is probably the major source of objects capable of causing
serious damage, because they can be of considerable mass.

The extent of the damage depends upon the size of the failed
component centrifuged into the steam path. Missiles could cause
massive damage on the first row of impacted elements and then be
chopped into a number of smaller objects, which would cause less
extensive damage (from a structural integrity perspective) as they
pass through the steam path. Figure 4.3.1 shows a control stage that
has suffered a blade failure at the root upper load bearing surfaces.
The blade has detached and become a missile within the unitthere
is relative little damage to the other blades in the row. Adjacent
blades need to be replaced, but other elements are, from a cursory
examination, in an acceptable condition. Nondestructive examina-
tion confirmed these blades had not suffered severe damage.

In this stage there was a large axial space between this row and
the following stationary blades inlet. Therefore, the detached blade

403
Turbine Steam Path Maintenance and RepairVolume One

was centrifuged into this steam chamber where it did not re-enter the
main steam flow.

By comparison, it is possible for a blade of much smaller mass to


fail and cause significant damage because there is no axial space for
it to rest in.

Fig. 4.3.1A failed blade in a control stage. There is


minor damage to adjacent blades and the cover band.

Massive objects can cause damage sufficient to force the unit


from service for extensive periods. If the rotor is thrown out of bal-
ance there can be extensive rubbing damage, which could be hard-
er to correct than the failure of a steam path component. Figure 4.3.2
shows the stationary blade row of a stage following a massive fail-
ure. Here the vanes have suffered impact damage, causing tears
adjacent to the outer diaphragm ring where objects have been pres-
ent in the axial gap between the rows.

404
Steam Path Damage Induced by Operational Events

Fig. 4.3.2Impact damage to a stationary blade row. The damage resulted from
the failure of a large rotating blade which was trapped in the axial gap
between the rows. A tear can be seen adjacent to the side wall.

Formed in the boiler


The boiler is a major source of small particles (i.e., weld spatter)
that are deposited on the tubing and eventually break loose and trav-
el to the turbine valving system. The screens mounted around the
valve inlet capture larger beads but smaller beads pass through the
screen and impact the steam path. Figure 4.3.3 shows a screen
removed from serviceweld beads can be seen. Figure 4.3.4 shows
typical damage resulting from small weld objects carried through a
steam strainer screen.

External to the steam path


For missiles to be generated externally and enter the steam path
is not common, but it does occur. If there is failure of any valve or
screen component, there is a high probability they will gain access

405
Turbine Steam Path Maintenance and RepairVolume One

to the steam path, and will impact on the blade elements if they can
pass through the first stage nozzles.

Fig. 4.3.3The screen removed from around the valves, showing metallic debris which
was too large to enter the steam path.

Figure 4.3.5 shows the result of the failure of a valve seat on the
control stage stationary blade row. Here a portion of the valve seat
was wedged between elements of the stationary blades, and a por-
tion managed to travel into the axial gap between the stationary and
rotating blade rows. The rotating elements are shown in Figure 4.3.6.
This damage was sufficient that smaller missiles were formed, trav-
eled downstream, and caused damage on a later stage (Fig. 4.3.7).
This impact damage became progressively less severe as the distance
from the control stage increased.

406
Steam Path Damage Induced by Operational Events

Fig. 4.3.4Peening type damage on a control stage stationary blade row. This damage
was caused by weld bead from the boiler superheater section and possibly other
small particles.

Fig. 4.3.5Damage to a control stage stationary blade row, caused by the


failure of a valve seat which then passed into the steam path and
become trapped between the fixed blade discharge and rotating blade
inlet. The nozzle partitions are damaged to the extent they must be
rebuilt by welding.

407
Turbine Steam Path Maintenance and RepairVolume One

Fig. 4.3.6The rotating blades from the control stage showing in figure 4.3.5. These
blades have suffered massive damage, and must be replaced.

Fig. 4.3.7Damage sustained further down the steam path by the stage showing in fig-
ures 4.3.5 and 4.3.6. Having been chopped into smaller particles, the valve seat passed
down the steam path causing subsequent impact damage.

408
Steam Path Damage Induced by Operational Events

Drop ins
Problems associated with metallic pieces left in or dropped into
a unit during an outage can be major ones to most users. Drop-ins
can be of any size, limited only by their ability to bypass the com-
ponents that prevent their being seen before the unit is closed up and
returned to service. Figure 4.3.8 shows remains of a bolt left in or
dropped into a unit during an outage and not removed. This materi-
al origin was traced from an analysis of the type of material.

Fig. 4.3.8Pieces of a stud or bolt dropped into the


steam path during a maintenance outage.

IMPACT DAMAGE
CLASSIFICATION
Unfortunately, it is impossible to classify this type of damage.
The question that always arises is, Should it be classified in terms
of the amount of area affected? In terms of the closure of the inlet or

409
Turbine Steam Path Maintenance and RepairVolume One

discharge edges? Or in terms of the most significant damage that


causes surface roughening, or causes small ruptures in the surface
of the blade?

The following discussion considers various types of damage and


its general form. However, in evaluating any situation, it must be
considered that a form of damage acceptable at one radial location
on one stage may be totally unacceptable at a different location on
another. Therefore, the following descriptions of damage are dis-
cussed, but no offer is made of the most significant.

What can be stated is that some forms or levels of deterioration


will obviously force the unit from service. Othersand those more
perplexing to the engineerask, When failure is not imminent,
what level of damage is acceptable or tolerable? These are levels
that contain not only uncertainties difficult to judge (in terms of their
effect on both efficiency and the probability of failure if corrective
action is not taken), but are also dependent upon many other more
complex factors:

How critical to grid system security is the unit on which dam-


age has occurred?

What is the potential cost penalty of continuing to operate


with the blades degraded?

What is the cost in terms of time and material of replacing or


repairing the damaged elements?

What is the probability of the elements failing before the next


planned outage?

These primary considerations are not considered exhaustive, and


the operator may well have to consider other factors that override
many of the economic requirements. However, if a situation could
lead to a massive failure, then safety becomes a consideration and
action must be taken.

410
Steam Path Damage Induced by Operational Events

Peening type damage


Peening type damage occurs when small pieces of debris that the
steam transports through the steam path collide with a blade surface.
The high relative velocities between the blades and the metallic
objects are sufficient to cause plastic deformation (impact craters) on
the blade surface. This form of damage occurs on both stationary and
rotating elements, indicating that the objects have themselves
achieved a high velocity within the steam.

Figure 4.4.1 shows the tip regions of a rotating blade row that has
suffered extensive peening type damage. There is also some small
amount of damage on the cover. This is not considered extensive, but
will have reduced the effectiveness of the integral seal produced on
the cover. This damage is consistent with the impact of small parti-
cles, formed in an earlier stage, that are trying to gain access to the
blade row. It can be seen that this damage does not extend too far
back into the passage.

The various sources of the small objects capable of causing this


peening type damage include the following:

When many fabrication (welding) processes used in the


manufacture of thermal cycle equipment are completed
equipment built in placeconsiderable quantities of weld
bead are often left in the components and pipe work con-
necting them

Many of these beads are attached by heat fusion to metallic


surfaces. Such fusion is not a metallic bond, and these
objects are capable of loosening during operation.

It is therefore common for such weld beads to be left inside the


boiler and other vessels and structures. Some are loose and some are
attached to the plates and tubes. Indeed, it would not be possible for
the manufacturer to remove this economically. During operation,
this bead will experience both steam pressure and temperature

411
Turbine Steam Path Maintenance and RepairVolume One

expansion forces causing it to detach from the surface to which it is


attached. These forces may however, not be of sufficient strength that
the bead detaches immediately upon start up. Under these circum-
stances, the unit can operate for some considerable period before
there is sufficient effect to loosen it. If the bead loses some of its
adhesion, it eventually detaches. The bead is then free to be carried
through components and pipe work into the turbine.

Fig. 4.4.1Peening type damage occurring on the inlet edge of a rotating blade row.
This level of damage can be tolerated from structural integrity considerations but will
have an adverse influence on stage efficiency.

It is common practice for the majority of manufacturers of steam


turbines to supply fine mesh screens for use during initial start up,
and use after any repairs requiring in-place welding that cannot be
cleaned at completion. These screens are used over the normal
coarse screen surrounding the main inlet valves. After a periodnor-
mally some weeks or monthsthese fine mesh screens are removed,

412
Steam Path Damage Induced by Operational Events

and any collected weld bead or other debris, is removed with it. At
this time the unit goes into its normal mode of operation, with only
the coarse screen in place.

This fine mesh screen cannot be tolerated in the unit indefinite-


ly because of the excessive pressure drop associated with its use and
the consequent loss in unit output and efficiency. Economics of plant
operation dictate this fine mesh screen should normally be removed
after about six weeks from initial start up. By that time, the majority
of the weld bead and other fine material entrained in the system
would have dislodged and been carried out. There is a possibility a
portion of this weld bead will continue to adhere to the equipment
surfaces for a greater period and then become loose after the fine
mesh screen has been removed. At that time, this bead is free to
enter the turbine steam path, with the possibility of causing peening
type damage.

Components that detach within the steam path will impact


with the rotating components immediately in their vicinity

These impacts chop (or break down) these detached pieces


into small objects of a size capable of causing damage. The generat-
ed or chopped objects are then freeand small enoughto pass
through the steam path and make impact craters in subsequent
stages. Figure 4.4.2 shows a stage that has suffered damage and shed
a portion of its coverband that was cut into smaller pieces and
caused peening-type damage on the following rows of the unit.

The rotating blades ability to chop depends upon stage temper-


ature. At the higher temperature, the detached components are more
easily broken down into smaller objects. However, the rotating blade
material is also softer at the higher temperatures, and may also suf-
fer more severe damage from the impacts.

413
Turbine Steam Path Maintenance and RepairVolume One

There is also the possibility that debris existing in other parts


of the system eventually may be able to migrate into the tur-
bine and be carried through the steam path

This debris can include small parts left in a unit at maintenance


outages and other foreign material that gains access through the per-
formance of work on the unit. For these reasons, it is normally nec-
essary for maintenance staff to be aware of any part or components
dropped into the unit during repairs or inspection. It is normally less
expensive to recover these parts than leave them in the unit.

Fig. 4.4.2A high pressure stage where a portion of the cover band has detached caus-
ing impact damage both to the blades in that row and to elements downstream.

When solid particle debris that exists in the turbineirrespective


of its method of entry or generationstrikes the blades, it will, upon
impact, carry sufficient force to damage the vanes. It can also impact
with the coverband, tenons, and any flow-restricting device (leakage
control seals) located above the blades. While this damage is usual-
ly in the form of small craters on the surface of the component, it can
cause a significant frictional loss.

414
Steam Path Damage Induced by Operational Events

There are no general rules as to the repair procedure, or when


they should be employed. Each case must be assessed individually,
based on:

the potential loss in efficiency

the possibility that such damage could deteriorate to the


extent the unit integrity is placed in jeopardy

Figure 4.4.3 shows peening damage on the inlet nose of a vane


with water collecting grooves. Such impact damage could effective-
ly remove the water-centrifuging capability of these blades, increas-
ing the moisture content of the steam path beyond the design-antic-
ipated level.

Fig. 4.4.3Minor impact damage to a vane inlet edge, but in this case the impacts are
threatening to block the moisture drainage channels.

Damage due to loose material in the steam path tends to be more


acute on the rotating blades, i.e., objects causing massive or heavy
craters on the rotating blades will cause peening type damage on the

415
Turbine Steam Path Maintenance and RepairVolume One

stationary blade elements. This is probably because there is a lower


velocity between the objects and the stationary blades so less dam-
age is likely to be caused there.

Deformed vane inlet edges


Large objects in the axial gap between the stationary and rotating
blade rows always carry the possibility of deforming and partially
closing the vane outlets. This is shown in Figure 4.4.4, where the
entire rotating blade row inlet has been damaged and partially
closed. This unit suffering high levels of vibration due to some phe-
nomenanot necessarily the failure that closed the blade inletsis
shown by the evidence of the heavy rubs on the coverband inlet side.

Fig. 4.4.4Impact damage on a vane inlet edge. These impacts are closing down in
inlet portion to the steam passage, restricting steam access to the row.

416
Steam Path Damage Induced by Operational Events

Large surface craters


Occasionally, a mechanical rupture occurring within the steam
path is not of a magnitude that the unit is immediately forced from
service. Indeed, many such events are not even detected, or
detectable, until the unit is removed from service for a maintenance
outage. These ruptures therefore go undetected, and the unit contin-
ues to operate for extensive periods of time with a damaged condi-
tion present. While such damage will cause a reduction in stage and
unit efficiency, this may not be detected unless the most sophisticat-
ed testing devices are used to measure section efficiency and unit
output.

Unless it escalates to an unacceptable degree by further ruptures,


this situation is acceptedoften in ignorance of its existenceand
unless blades in other stage elements have been damaged as a con-

Fig. 4.4.5Large impact craters on a rotating blade inlet edges. These craters have
occurred in a region where tears can easily develop. This situation should be correct-
ed before the unit is returned to service.

417
Turbine Steam Path Maintenance and RepairVolume One

sequence of the situation, this is of no great concern to the operator.


However, when this condition is discovered upon opening the unit,
an evaluation must be made and the situation corrected if further
damage could result.

This type of situation is shown in Figure 4.4.5. Large craters and


tears exist on the inlet nose of a blade row, and many of these
indents, craters, and tears could, if not corrected, eventually force the
unit from service. The decision to decline corrective action and return
this unit to service can only be taken recognizing the risk involved. A
similar failure is shown in Figure 4.4.6. After separation, the detached
main objects are chopped into smaller objects. A portion of cover-
band from the L-1 stage has detached after a high-cycle fatigue fail-
ure of the tenon. Such a failure would probably not be noticed
there might be a marginal increase in the level of vibration on the
bearing closest to the stagebut this might not draw attention to the
failure. However, the consequence of this coverband detachment can

Fig. 4.4.6A missing segment of cover band from the penultimate stage of a low pres-
sure expansion.

418
Steam Path Damage Induced by Operational Events

be seen in Figure 4.4.7. There has been damage caused to the inlet
edge of the following (last or exhaust) stage rotating blade. This does
not appear to be significant at this time but there is always the possi-
bility of a crack propagating at some later date.

Fig. 4.4.7Impact damage on a last stage blade,


where the cover banding of figure 4.3.6 has impact-
ed in the erosion shield region.

When minor ruptures occur, they represent the detachment of


parts of the blade system and the generation of loose material within
the steam path. This material is then free to be chopped and rebound-
ed between the components, where it normally will be cut into
objects small enough to pass through the blade openings with the

419
Turbine Steam Path Maintenance and RepairVolume One

main steam flow. During their passage through the steam path, this
debris impacts on stationary and rotating vanes, damaging or deform-
ing them, and causing an efficiency losspossibly producing signifi-
cant mechanical damage. Figure 4.4.8 shows the damage caused to a
sealing strip attached to a stationary blade row outer ring. This type of
damage can represent a significant loss in unit output if it becomes
excessive. Impacts on the face of the outer ring also are seen.

Fig. 4.4.8A damaged radial seal strip. Because of the


thin section, seal strips are particularly susceptible to
impact type damage.

Massive craters and damage


Any component failure (mechanical rupture) occurring within
the steam path has the potential to generate massive metallic debris.
This debris will often be chopped by the rotating blades into smaller
pieces, and pass through the remaining downstream portion of the
steam path, probably causing minor mechanical damage as dis-
cussed previously. Due to this chopping action, and during passage
through the blading system, these initially large pieces can cause
severe impact damage to the components, depending principally
upon their size, hardness, and relative velocity at impact.

420
Steam Path Damage Induced by Operational Events

While the rotating blades are most susceptible, the stationary


blades can also suffer massive damage. This is shown in Figure 4.4.9
for a stationary blade row and in Figure 4.4.5 for a rotating blade.
Another example of this form of major damage on a high-pressure
stationary blade row is depicted in Figure 4.4.10, where portions of
the discharge edge are missing, and large sharp-edged craters exist
on the suction face. Figure 4.4.11 shows another form of this dam-
age on rotating blades. Complete discharge edges have broken loose
as the result of impacts from free materialin this case, generated
within the row itself.

Fig. 4.4.9A stationary blade row with very large craters at the discharge
edge.

Material ruptures
Depending upon the point and angle of impact, and the size of
the particles that impact on the vanes, tears can be formed. This is
most likely to occur when the impact is on the thinner inlet nose of
the rotating blade, or when debris is trapped between the discharge
from the stationary and inlet to the rotating rows.

A tear is shown on the discharge edge of a stationary blade row


(Fig. 4.3.2), and in a rotating blade row (Fig. 4.4.5). Objects trapped
between the rows caused both tears. A similar tear, in which there

421
Turbine Steam Path Maintenance and RepairVolume One

are multiple impact craters on the stationary blade vane discharge


edge, and a tear on one element at the outer sidewall is shown (Fig.
4.4.12). Figure 4.4.13 shows a similar situation on a rotating blade
row, where a heavy impact has caused a tear to appear on the dis-
charge edge. Impact craters can also be seen on the coverband.

Fig. 4.4.10Massive type craters on the discharge tail of a fixed blade row.

Fig. 4.4.11High pressure blades having been impacted on their inlet edge,
have suffered material rupture.

422
Steam Path Damage Induced by Operational Events

Fig. 4.4.12A stationary blade with material ruptures at the discharge


edge.

Fig. 4.4.13A rotating blade row damaged by large particles


trapped between the blade rows.

423
Turbine Steam Path Maintenance and RepairVolume One

Elements trapped between the rows


Major damage normally occurs when an object too large to pass
downstream becomes trapped between the rows. It is not wedged in
a position where it will cause damage and is free to rebound and be
chopped into smaller pieces before passing downstream. Aspects of
this situation can be both advantageous and disadvantageous:

While in the inter-stage position (unable to migrate down-


stream), the object will not damage the other blade rows. If
this situation occurs and the piece can be removed, the dam-
age will remain limited to two blade rows

Stationary blades can normally be repaired, returned to serv-


ice, and continue to operate in an as new condition. There
could be the necessity of an extended outage to make such
weld repairs

Rotating blades suffering damage of this type most often


require complete replacement. This will normally require an
extended outage

Figure 4.4.13 is a rotating blade row that has suffered extensive


damage. The elements must be replaced before the unit can return to
service. In this case the damage extends to the coverbands and the
tenon heads, which have been completely destroyed by the rubbing
of objects between the coverband and casing. A similar situation is
shown in Figure 4.4.14, where objects have become trapped above
the coverband, causing serious gouging damage. The outlines of the
tenon heads can still be seen. There is deep grooving where the
debris has gouged the coverband outer surface.

Vane damage and assessment for correction


When impacts occur on a vane, and it sustains damage that
deforms the profile, at what point is it necessary to replace, refurbish,

424
Steam Path Damage Induced by Operational Events

or repair the affected elements? This is a major consideration when-


ever impact damage is found.

We know if any profile deformation has occurred, blade effi-


ciency will deteriorate from its previous condition. Depending upon
the nature of the damage, deformation of one profile can modify the
shape and expansion efficiency of two passages.

Fig. 4.4.14Grooves in a cover band as a result of debris trapped


between the cover band and the casing.

In general, peening type damage can be dressed. While not


restoring efficiency to the original level, it can at least minimize loss-
es. The following line of corrective action is suggested:

For stationary blade elements. If the extent of damage is at or


beyond the level of that shown in Figure 4.3.4, corrective action

425
Turbine Steam Path Maintenance and RepairVolume One

must be considered. However, unless there are significant tears, or


situations indicating objects will detach after return to service, the
economicsfuel cost and load factors against repair costsmay
also influence the final decision.

For rotating blade elements. Rotating blades are more sensitive


to damage. While peening type damage can be dressed to reduce
efficiency losses, no decisions should be taken that will place the
security of the unit in jeopardy. Apparently insignificant damage may
quickly deteriorate once the unit is returned to service.

While a user can often accept efficiency deterioration and con-


tinue to operate with damage, few situations arise where a unit can
be operated at risk. To do this, the situation must be fully analyzed.
It must be known that deterioration can quickly occur with the con-
sequences fully understood.

SOLID-PARTICLE EROSION
(ABRASION)
Todays fossil-fueled turbines work under higher pressures and
temperature conditions and are expected to endure more frequent
starts and shutdowns. This has given rise to solid-particle erosiona
damaging phenomena causing significant material loss from the var-
ious components in the stages following the initial entry or reentry of
hot steam from the boiler superheater or reheat section.

Damage is introduced by the action of hard scale that forms on


the water side of the boiler tubes. It is carried over into the turbine
steam path, where it contacts the various surfaces of components
and removes material from them. In such instances, the first one or
two stages of the unit suffer a recognized pattern of material loss in

426
Steam Path Damage Induced by Operational Events

various regions after the entry of steam. This damage can be severe.
Material loss occurring in regions which can affect both the efficien-
cy and structural integrity of the unit.

This damage occurs principally on the pressure faces of the sta-


tionary and rotating blade vanes, at the inlet and discharge edges.
However, such damage is also found as a material loss on the cover-
band and tenon heads.

These damaging effects have been determined to be the conse-


quence of hard particles of scale, which have been exfoliated from
the inside water surface of the boiler tubes. This scale is carried in
the form of small particles into the first stages of the unit, where they
come into high velocity contact with the stationary and rotating
blades. The scale is composed primarily of magnetite, which is
extremely hard, and initially has sharp or jagged edges. These dam-
aging phenomena can affect the stages of the high and reheat sec-
tions of the unit. It is believed there are two basic mechanisms con-
tributing to this damage:

A gouging or cutting action that occurs on ductile blade


materials. This is most severe when the impact angle is from
20 to 40. These damage incidents for oblique impacts are
a function of particle velocity raised to a power n, where
n has been reported to have values of from 2.0 to 4.0

A chipping type mechanism, in which small particles are


broken from trailing edges. This is a brittle type failure found
most commonly in austenitic materials

Material removal mechanisms can therefore be either abrasive or


have an impact effect on the scaleor a combination of both.
However, any removal mechanism must take into account material
losses on the surfaces being eroded, and how they react to scaling at
the high temperatures at which the steel operates.

427
Turbine Steam Path Maintenance and RepairVolume One

As scale enters the steam path, the steam will transport it through
the first stage nozzles. After its initial impact, a portion of the scale
will be centrifuged to the casing, where it may take a number of
routes through the remaining stages of the expansion, in some cases
colliding with and damaging other portions of the steam path. Other
portions of this scale will rebound in the axial gap between the sta-
tionary and rotating blade rows, eventually passing through the rotat-
ing blades and causing some level of damage to them. Eventually all
of this scale will migrate to the blade outer flow sections. Continuous
impacts with the blades, and the centrifugal force imparted to the
scale, means the scale will be deposited at the casing into various
niches, where it will remain.

The scale is formed in the boiler tubes as a result of continued


operation at high temperatures, and is attached to the tubes. However,
as temperatures change within the tubesparticularly at start up and
shutdownscale will exfoliate and become free within the tubes
themselves. From this observation it is clear that units on cycling duty,
or subject to a large number of cold starts, will suffer more exposure
to this form of damage. They will require a greater level of monitoring
and ultimately require greater levels of maintenance.

SCALE FORMATION
Boiler tubes are normally produced from a low-alloy steel (per-
lite), which lends itself to relatively easy oxide scale formation on the
internal surfaces.

Figure 4.6.1 shows magnetite scale on the steam side of the boil-
er tube. During startup and shutdown, when temperatures within the
boiler and reheater tubes are changingor when temperature tran-
sients occur in the vessels due to uneven firing or other conditions

428
Steam Path Damage Induced by Operational Events

both the tube and the scale expand and contract. Because the tube
and scale have different coefficients of expansion, shear forces
develop along the interface with temperature changes. These forces
are of sufficient magnitude that with large temperature changes, they
will cause rupture at this interface, allowing scale to detach.

Fig. 4.6.1Magnetite formation on the steam/tube interface.

Increased flow velocities through the boiler tubes can help


loosen and detach the oxide scale. When this scale breaks loose, it
is carried over into the steam turbine stages, damaging the compo-
nents with which it comes in contact. Figure 4.6.2 shows the stem
from a stop valve pilot that has been exposed to scale impact for two
years; Figure 4.6.3 shows the valve seat from this same valve. The
material loss due to impingement by the scale can be seen. After
passing through the valves, the scale enters the main steam path.

429
Turbine Steam Path Maintenance and RepairVolume One

Fig. 4.6.2The stem of a pilot valve which has been


subjected to solid particle erosion.

Fig. 4.6.3The seat of an eroded pilot valve.

430
Steam Path Damage Induced by Operational Events

THE EROSION MECHANISMS


In considering the mechanics of the cutting or abrasive actions that
remove material from the stationary and rotating rows, it is best to
review typical stage velocity diagrams, and consider the probable track
of the particles between and within the steam expansion passages.

Actual material losses depend mainly upon stage geometry and


the turning angles of the vanes. These losses are therefore design sen-
sitive. However, general explanations of the mechanics of the two
types of row (stationary and rotating) can be developed, and with
observation of individual units, a better understanding of the indi-
vidual loss modes developed.

The stationary blade row


Consider the velocity profiles of the stationary blade row shown
in Figure 4.7.1. In such a blade row, steam carrying small particles
of scale enters the blade passage. Because steam particles within this
passage are of smaller mass, they can be accelerated by the release
of thermal energy deflected through the vane-turning angle. Both the
pressure surface of vane R and the suction face of the adjacent
vane S influence them. At the throat discharge, the steam leaves
the row at the design angle.

However, because the transported scale particles are of greater


mass, they cannot be accelerated through the complete passage
turning angle while being deflected to some degree through the pas-
sage. Some of them make impact contact with the pressure face of
the stationary vane R, causing a cutting or abrasive action on its
surface. That scale contacting the vane at point j will be predomi-
nantly impacting, while the scale contacting at k will be predom-
inantly gouging. However, every piece of contacting scale will both
impact and gouge. It is a matter of extent, and that extent varies from

431
Turbine Steam Path Maintenance and RepairVolume One

j to k, with the actual damaging mechanism being dependent


upon the impact angle.

Fig. 4.7.1The steam and scale tracks between the blades, also show-
ing the potential wear pattern at the discharge edge.

When it contacts the pressure surface of vane R, the scale will


react on it and remove material. This material removal will continue
to penetrate, removing more surface material towards the discharge
edge because of the greater quantity of scale present along the sur-
face of the vane. Shown in more detail in Figure 4.7.1 is the effect of
subsequent levels of penetration z. The amount of vane material
remaining after these subsequent levels of penetration will provide
only small visual changes to casual inspection up to the position
where suction face material is removed. At this point, the more char-
acteristic loss patterns occur.

The rotating blade row


The tracks of the scale in the rotating blade row are influenced
by the forces exerted on them by the flowing steam and by the cen-

432
Steam Path Damage Induced by Operational Events

trifugal effects of rotation. This means that for the rotating rows, it is
necessary to consider the scale moving in the axial/tangential plane
and also the radial direction. Whats more, in the rotating blade
rows, the mechanism of material loss is somewhat more complex,
having the complication of high radial flow components introduced
by the centrifugal forces, imposed on the drag forces acting on the
moving scale by the main steam flow.

The axial/tangential planes. Consider velocity triangles in the


axial/tangential plane for the steam and scale in a typical rotating
row in Figure 4.7.2. Here the velocity triangles of both steam and
scale entering and leaving the rotating blade row can be seen.

C1e
c
c b
b a
a U
W1e
W1s
C1s

W2e c
b c
a b
U a C2s
W2s
C2e

Figure
Fig. 4.7.2The velocity triangles 4.7.2 and scale at discharge from
for steam
the stationary bladetriangles
The velocity row. for steam and scale at discharge from
the stationary blade row.

433
Turbine Steam Path Maintenance and RepairVolume One

As shown in this vector diagram, the rotating vane, having a tan-


gential velocity U, is arranged to accept the inflowing steam from
the stationary row with a relative velocity W1s. This steam flows
into the expansion passage between the blade vanes and discharges
(with a substantially axial velocity) into the following stationary
blade row. The particles of scale have essentially the same discharge
angle as the steam leaving the stationary blade row. It should be rec-
ognized that there could be discharge angle differences for that por-
tion of the scale contacting the stationary vane discharge edge. Such
scale could rebound or deflect from the steam angle. The relatively
larger mass of the particles of scale will leave the influence of the sta-
tionary blade row at lower velocities C1e.

In the vector diagram (Fig. 4.7.2) velocity triangles for three dif-
ferent particles (a, b, and c with velocities W1e) are displayed.
These different velocity triangles reflect a possible range of discharge
velocities for the scale, with actual velocities influenced by their
size, shape, and history of previous impacts. Therefore, the scale
entering the rotating blade row can be conveniently divided into two
groups:

Scale that has passed through the stationary blade row with-
out collision with the stationary vanes. It will tend to be of
smaller size, and have somewhat higher velocities, since it
has not lost energy by impact

Scale that has impacted on the stationary vanes. It will have


lost energy, and therefore have a lower velocity, but might
also have been broken into smaller pieces by the effect of
impact

It is impossible to take all these factors into account in develop-


ing a detailed model. It is sufficient to assume the discharge veloci-
ties from the stationary blade row can have a spectrum of values
from W1e(a) to W1e(c) as shown for the inlet velocity triangles.

434
Steam Path Damage Induced by Operational Events

It is now necessary to consider the possible trajectories of scale


particles at both ends of the probable velocity spectrum. Consider
first that scale entering with a velocity W1e(a) (the scale that has
passed through the stationary blade vanes without impact, and will
possibly be of smaller size). This scale will enter as shown in Figure
4.7.2striking the rotating vane suction face at position M. At
impact, this particle will possibly fracture into smaller pieces and
rebound into the main steam flow. It is then probable all or most of
this scale (or smaller particles) will be carried through the steam pas-
sage to enter the following stationary blade row. Scale particles
entering the rotating blade row with a velocity W1e(c) will strike
the vane at position N and rebound. Because the majority of this
scale has already impacted the stationary blade vanes, it will tend to
be of larger size, and the majority will probably rebound into the
axial gap between the stationary and rotating rows. This scale will be
considered in more detail in the following section, which discusses
the radial flow effect.

The scale carried through the expansion passage to discharge


will leave the rotating blade row with velocities and directions as
indicated as C2e, which again can have a spectrum of values of
both velocity and discharge angles. The steam will leave the row at
the design predicted velocity C2s in a substantially axial direction.

The radial plane. Due to the centrifugal effect on the working


fluid in the rotating blade passages, scale particles flowing in the
passage formed between these vanes will experience considerable
radial forces that will tend to make them flow to the outer radii of the
row, where any coverbands will prevent them being centrifuged
radially outwards onto the casing. Figure 4.7.3 shows scale paths
indicating the radial flow towards the underside of the coverband,
though the coverband prevents these particles from leaving the row.
However, upon leaving the blade row, these particles will have a
large radial flow component at its discharge edge due to their mass.
This will cause a large proportion of these particles to flow onto the

435
Turbine Steam Path Maintenance and RepairVolume One

casing. Some of this scale will rebound from the casing and re-enter
the main steam flow. A portion will be retained there in any form of
hideout or cavity that exists. Often when a unit is opened, some
scale can be recovered, principally from the lower-half casing. A
probable scale flow pattern in the rotating blade is shown in Figures
4.7.3 and 4.7.4.Here the effects of steam leaking over the coverband
are also shown.

Ca

Cr

Cg

Fig. 4.7.3The radial flow4.7.3


Figure of scale particles
after
Theleaving the stationary
radial flow blade row.
of scale particles after leaving
the stationary blade row.

The amount of steam leaking over the coverband has a direct


effect on the quantity of scale reaching the casing through any
sealsaxial or radialthat may be designed into the stage. Figure
4.7.4 shows the effects of leakage and the interference that can
occur above the blade tip. This figure also shows the possible paths
the scale can take. After initially passing to the casing on the rotat-
ing blade inlet side, it can impact on the casing and will swirl in the
radial clearance above the blade. This scale will eventually be car-
ried past the radial seal and interact with the steam discharging from
the rotating blade row. It can possibly be influenced, to a lesser
degree, by the scale centrifuged to the casing from the rotating
blade. Another consideration in terms of the seal influence is that as

436
Steam Path Damage Induced by Operational Events

seals rub, clearances will be opened, producing an increase in the


leakage steam quantities and steam forces affecting the scale parti-
cles. Greater steam forces tend to effectively suppress scale radial
flow and permit less scale centrifuged to the casing. Scale that fails
to reach the casing and is retained in the steam flow will be carried
into the following stage.

Scale carried over


with the steam

Swirl

Scale rebounding
from the casing

Fig. 4.7.4Scale rebounding


Figure 4.7.4from the casing inner
surface and moving
Scale rebounding past
from thethe seal strip.
casing inner surface and
moving past the seal strip.

Consider now in more detail possible paths of the two


extremes of particles a and c (Fig. 4.7.2). The particles indicat-
ed by condition a with a velocity W1e(a) enter the steam path at
a high velocity and penetrate the expansion passage to the point M
before they impact on the suction face. The majority of these parti-
cles are transported through the blade passage. While they will be
centrifuged radially outwards to the casing, a proportion of them will
flow into the following stationary blade row, possibly after impacting
on the casing inner surface. This centrifugal action within the blade
passage is shown is Figure 4.7.3.

437
Turbine Steam Path Maintenance and RepairVolume One

Of greater concern in terms of material loss and damage, is that


scale (particles c entering with velocity W1e(c)) rebound into the
axial gap Cg between the stationary and rotating rows (Fig. 4.7.3).
This scale, causing some level of damage to the vane with its initial
impact and upon rebounding some distance back into the axial gap
Cg, will be re-entrained into the main steam flow and be free to re-
enter the rotating blade row causing further damage.

Upon leaving the vane after impact at point N, this scale will
have some small radial component. Upon entering the axial gap, it
will tend to flow or migrate radially outward with some velocity.
Depending upon the size and direction of the scale, it will flow back
and strike the stationary blade row, where it will rebound and be car-
ried by the steam field force back into the rotating blade row. This
type of impact and the consequent damage can be repeated as many
times as the particles are forced by the steam back into the blade
expansion passage. Figure 4.7.5 shows the possible path of a single
piece of scale that enters the rotating blade row near the root sec-
tion. Its multiple impacts in the flow ultimately lead to the tip sec-
tion, where it will either pass over the coverband or through the
blade row. It is also possible that any given piece of scale will frac-
ture into smaller pieces developing individual paths appropriate to
their size and relationship with the steam field. Similarly, scale enter-
ing at other locations will rebound and flow towards the tip section,
so there is a gradual accumulation of scale towards the blade tip. It
would be expected that damage would become progressively more
severe towards the tip.

At unit startup, the quantity of steam and the force of the steam
being admitted to the unit is relatively small. This is when the major-
ity of scale exfoliates and flows to the turbine. However, there could
be some delay in the entry of this scale to the steam pathparticles
may be too large to pass through the fine mesh screen, become
trapped in pockets and hideouts in the valves and inlet pipes, require
some time to break into pieces small enough to pass through the

438
Steam Path Damage Induced by Operational Events

screen, and other scale may eventually be torn from hideouts and
pass to the turbine. Therefore, while scale is carried to the turbine at
start up, its entry to the turbine in a form causing damage may be
delayed by various mechanisms, and the damage process could con-
tinue for some time after each return from service.

Ca

Cg

Fig. 4.7.5Scale rebounding


Figure 4.7.5 between
the stationary and rotating
Scale rebounding blade ele-
between the
ments and migrating
stationary out towards
and rotating the tip
blade elements
section
andand seal. out towards the tip
migrating
section and seal.

MATERIAL LOSS PATTERNS


DUE TO SPE
Material loss patterns that occur on stationary and rotating blade
rows are now familiar to the majority of operators, and immediately
recognized. Such damage is characterized by the modification it
produces to the cross section on both sets of vanes and sidewalls,

439
Next Page

Turbine Steam Path Maintenance and RepairVolume One

and the effect it has on the sealing system above the rotating blades.
Removing this material therefore modifies the shape of the expan-
sion passages, and is capable of adversely influencing the steam path
expansion efficiency and the unit reliability.

These loss patterns from the stationary and rotating rows have
certain similarities, but the form the damage within the two rows
takes can also be quite distinct. This is due to different forces influ-
encing the scale paths as it flows though the rows. It is of value to
consider the two rows, and the forms of damage to which both are
subjected.

The scales hardness and shape can cause erosive damage before
its entry to the steam path, and it is necessary for the operator to
examine the valve control system to ensure the level of damage
these components sustain is within acceptable levels.

Early turbine damageafter initial admission or after return of


steam from the boiler reheater sectioncan be subject to relatively
severe damage as a consequence of this scale. Such solid-particle
erosion of the unit is normally a slow process, although its effects
and presence can be inferred from both pressure distributions
throughout the unit and state line efficiency measurements. This is
because a gradual process must be monitored with test-quality
instrumentation. It is often difficult, immediately after start up, to dif-
ferentiate between the effects of this erosive damage and damage
due to impurity deposition on the steam path expansion surfaces.

Removal of material occurs primarily at start up, and if steam is


admitted to the steam path through partial arcs of a row, it is common
for those segments which are first to admit steam to suffer extreme
damage. This can be seen from an examination of a control stage
nozzle box. Figure 4.8.1 shows a portion of a control stage nozzle
plate, in which material loss is not even around the circumference.
This row has suffered material loss from the stationary blade vanes on
their discharge edge; the uneven nature of the erosion damage is also

440
Chapter

5
Steam Path Damage
and Deterioration
from Material
Property Degradation

INTRODUCTION
Within the turbine steam path and for major components of the
turbine, many forms of both static and dynamic loading introduce
mechanical stress. In addition, many components that operate at
high stress levels do so continuously at elevated temperatures, which
while they are sensibly constant, will under certain conditions
change at rates capable of causing changes within the material.

Under such conditions of load, temperature, and temperature


variation, these components are, or can be subject to various forms
of material property degradation and material life consumption. The
design process evaluates these conditions and attempts to define the
individual components so they are suitable for operation with the

487
Turbine Steam Path Maintenance and RepairVolume One

stresses induced by the operating conditions, enabling them to


endure for the anticipated life of the unit.

However, these phenomena are capable of damage to the extent


that component failure occurs. Circumstances can occur during
operation where the unit is operated beyond its design limits. These
circumstances represent conditions under which stresses are
induced beyond those anticipated. Similarly, material properties
inferior to the design specification can be introduced. While failure
may not be instantaneous, there is an accumulation of damage that
ultimately forces the unit from service and requires corrective action,
either immediately or for a future outage.

Steam turbines are intended and designed to operate for consid-


erable periodspossibly between six and eight years between major
inspection outages. Therefore, the internal portions of the unit,
including the steam path hot sections will not be available for
damage assessment and correction between these periods. It is nor-
mally considered necessary that clearances and deflections are
measured for comparison to earlier measured values recorded when
the steam path became available at a previous outage (see chapter
1). These inspections and comparisons are necessary to ensure pend-
ing problems are evaluated and, if necessary, corrective action taken
to avoid returning the unit to service with the risk of a forced outage.

Major material degradation phenomena that can potentially


affect the steam path components include high-temperature creep
and high- and low-cycle fatigue. These three phenomena will be
considered in this chapter.

488
Steam Path Damage Induced by Material Degradation

CONSIDERATIONS OF
MATERIAL STRUCTURE
As a molten metal transforms from liquid to solid state, a random
distribution of the nuclei form a eutectic. These nuclei then grow by
the addition of other atoms of the pure metal in a closely defined and
repeated matrix. The number of nuclei that form within the melt
and the size of the individual grains that grow and form the final
solidis influenced by the rate at which the melt cools. With a slow-
er cooling process there is greater opportunity for the individual
nuclei to initiate and crystals to form.

Because of uneven cooling within the melt, the initial formation


of the nuclei is most often towards its outer boundaries, where the
temperature first begins to lower by heat loss. This cooling, and the
subsequent solidification of the melt continue, with individual crys-
tals of the alloy growing until the liquid phase has been completely
transformed to the solid. This formation of the nuclei is random, as
are the growth and orientation of the individual crystals.

Because of this random nature, crystal alignment is such that


individual crystals will not join. Instead, the solid will exist contain-
ing the crystals of the alloy, with boundaries filling the space
between the individual crystals. At the boundaries of the individual
metal crystalswith each crystal representing a correct spatial
arrangement of the individual metallic atomsthere are discontinu-
ities. These discontinuities are the crystal or grain boundaries and
represent a structural discontinuity that can contain impurities of ele-
ments and compounds that were present, but not absorbed by the
individual crystals. They may also contain trace amounts of undesir-
able elements unable to be removed from the original melt. This
boundary material can be either enriched or depleted of one of the
alloying elements, and therefore capable of influencing the total
physical properties of the material.

489
Turbine Steam Path Maintenance and RepairVolume One

Under development are materials for turbine application that are


of a single crystalsome success has been achieved. However, the
majority of materials is, and for a long time will continue to be those
with a structure of individual crystals connected at their grain
boundaries.

The most common melt used within the turbine materials has
two distinct material characteristicsthose of the pure metallic crys-
tals and those of the material contained within the boundaries. For
materials produced by modern methods (with controlled con-
stituents and cooling rates), the characteristics of the grain boundary
material can be predicted and even controlled. These boundary
materials offer properties fully acceptable, and even advantageous to
the total properties.

Note: Single crystal blades that have been developed are cur-
rently used in gas turbines where the operating temperature is high-
er. These blades can operate at higher temperatures and have a creep
life greatly in excess of the more common alloy steel used for steam
turbine elements.

HIGH-TEMPERATURE CREEP
Creep is a material phenomenon affecting the high-temperature
regions of the steam path and those components exposed to a ten-
sile or shear stress. More accurately, creep is defined as high-tem-
perature creep, because while materials under stress creep at all
temperatures, it is only at higher temperatures that these deflec-
tionsand ultimately material ruptures due to stressbecome sig-
nificant. It is necessary to recognize that the higher the component
temperature, the faster the creep deflection, and the faster the final
failure will occur.

490
Steam Path Damage Induced by Material Degradation

The design process is aware of, and allows for, creep deformation
in those components and regions where it can occur. If the conditions
exist sufficiently to produce creep in a component, then that compo-
nent will suffer the effects of creep deformation. What the design
process will do in such a situation is determine the material, dimen-
sions, and other necessary characteristics so the component will
deform by an amount that can be tolerated for the operating life of the
unit. However, it is necessary once again to consider that the unit may
operate beyond its design tolerances. Under that situation, creep rates
and total deflection could occur that are beyond an acceptable limit.

For a component that is affected by creep, and for the distortion


to become significant, it is necessary to have present a stresseither
uni- or multi-directionaland for this stress to be applied consis-
tently over extensive periods of operation. The creep strain that
results can occur at varying rates, and is normally, by definition,
divided into primary, secondary, and tertiary phases. This means that
creep rates are not constant, but vary with time. The primary phase
is rapid, but as the creep rate enters the secondary phase the defor-
mation slows to an almost constant value before accelerating to fail-
ure in the tertiary phase. However, in most cases within the steam
turbine, the presence of creep is noted well before the damage
extends to the extent of mechanical rupture.

For creep deformation to be considered significant within the tur-


bine steam path, it is not necessary for mechanical rupture to have
occurred, or even for cracks to have formed. This is because at many
locations within the unit there are relatively small running clearances
between stationary and rotating parts. Should there be rubs due to
creep deformation, then significant damage can result with major
components being involved. While rubs should not occur in the pri-
mary phase of creep deformation, and deformation of a well-designed
component should never enter the tertiary phase, it remains the
responsibility of the design engineer to select component geometry so

491
Turbine Steam Path Maintenance and RepairVolume One

deformation during the secondary phase will not cause such rubbing
contact to occur.

The creep rate for any component, under a defined set of oper-
ating conditions, is dependent upon two major parametersoperat-
ing stress level and operating temperature.

The operating stress level. The stress levels experienced by any


component is dependent upon a number of variables, including
steam pressure drops, rotational speeds, and steam flow rates. Each
of these parameters are capable of being varied during operation,
and therefore can modify local stress levels. The stress developed on
any component or part of the unit may also vary during operation.

A consideration with the transient variation of these operating


parameters is that they may cause a local temporary (elastic) defor-
mation. While not permanent, it may cause a total instantaneous
deflection or distortion during the transient, such that mechanical
damage could be initiated. This failure may not have occurred if the
existing creep (plastic) deformation had not been so severe at the
time of the transient.

The operating temperature. Component temperature as stress is


applied influences the materials ability to resist the applied stresses,
and therefore control the rate at which creep deformations occur. As
the temperature of the component rises, creep rate increases. For this
reason, it is necessary to control boiler superheater and reheater
temperatures, and control them within design specified levels. These
design- imposed restrictions, while appearing stringent, are intended
to extend the life of the individual components and by implication,
that of the unit.

As in the case of stress, considerations of temperature transients


are important, as these can represent a temporary but possibly dan-
gerous variation from the design specified operating mode.

492
Steam Path Damage Induced by Material Degradation

While the design process recognizes, attempts to anticipate, and


then allows for the possibility of transients during the normal oper-
ating life of the turbine unit, their accumulated effects can still result
in component damage or (in the most severe case) failure, with the
unit in a forced outage status. The design process allows for the
effects of transients by determining stress levels, and applying what
experience has indicated is an adequate factor of safety. However, it
is ultimately the responsibility of the operating engineer to limit, as
far as practical, the occurrence of transients that can reduce the
operating life of the unit.

CREEP DEFORMATION
Creep is a slow plastic deformation (considered briefly in chap-
ter 2). It occurs in any component under stress for extended periods.
This stress can be at a steady level or fluctuate, but the stress must be
in the same direction at all times to cause deformation. The stresses
required to cause this material deformation are much lower than the
values required to cause failure when applied suddenly.

The rate at which deformation under conditions of creep occur


is defined as occurring in three separate phases. While not entirely
distinct in any material, they can be recognized from a creep
strain/time characteristic curve (Fig. 5.4.1). This curve is construct-
ed with the assumption that both stress and temperature were held
constant during the test period.

Note: Creep deformation is a plastic flow of the material and is


not recovered when the load is removed. This differs from deflection,
which is elastic and recovered as load is removed, with the compo-
nent returning to its original form.

493
Turbine Steam Path Maintenance and RepairVolume One

Figure 5.4.1The phases of creep deformation to mechanical rupture.

The three separate phases to creep deformation can be explained


as follows:

Primary phase, A-B. During this initial phase, relatively


rapid extension occurs with application of the same constant
load, normally at a slightly decreasing rate. The rate of defor-
mation is initially rapid because the weaker locations within
the boundary layers fail at relatively smaller applied loads

Secondary phase, B-C. After this initial deformation (pri-


mary phase A-B), the creep rate decreases to a relatively
constant rate. After the initial boundary layer failures there is
a continuing series of ruptures as more boundary cracks
occur and fewer metallic crystals carry the load. During this
time the deformation continues, but there is little or no accel-
eration of the rate

Tertiary phase, C-D. As the deformation enters the tertiary


phase, the rate of extension accelerates until condition D
is reached, at which time mechanical rupture occurs. At con-

494
Steam Path Damage Induced by Material Degradation

dition D the material is no longer able to carry the applied


load due to the boundary layer ruptures.

Characteristic curve families


Consider a family of creep curves for a component material
under constant load but at different temperatures. A series of char-
acteristics (Fig. 5.4.2) is obtained. In these four curvesT1> T2> T3>
T4it as temperature rises so does the creep or deformation rate.
The following observations can be made concerning this family of
temperature characteristics:

At the higher temperatures, the creep rate is higher in each


of the three phases

At all temperatures, the amount of time in the primary phase


is about the same, but the total deformation is somewhat
greater at the higher temperatures

The time in the secondary phase is reduced as the tempera-


ture is increased, so that the creep deformation enters the ter-
tiary phase at a lower total deformation

The total life of the component is reduced at the higher oper-


ating temperatures

A similar family of curves can be constructed for a material held


at a constant temperature and tested at various stress levels.

As the characteristic curves leave one phase and enter the next,
the exact form of the curve is difficult to define. It is best described
as a change in the rate of creep deflection. Because the curves are
the result of experimental data, they may not be precise but provide
the design engineer with sufficient information that life predictions
can be made with a reasonable degree of accuracy.

495
Turbine Steam Path Maintenance and RepairVolume One

Tertiary
phase Mechanical
rupture
Extension " "

Secondary
phase
Primary
phase
Entrance to
Tertiary
T1 phase
T2 T3
Entrance to T4
Secondary
phase

Time

Figure 5.4.2The effect of temperature


Figure on stress rupture characteristics at
5.4.2
constant stress
The effect levels.
of temperature on stress rupture characteristics at constant stress
levels.

In defining the required creep properties of any material used for


a steam turbine component, it is necessary to recognize that coarse
grain materials creep at a lower rate than those with a fine grain
structure. Unfortunately, a component that is to be subjected to
creep-type loading may be produced from a fine-grain material
allowing it to meet other requirements of design. To a degree, this
compromises the creep characteristics of the material from which
the component is constructed. These considerations require a
detailed review and material evaluation. This is an evaluation the
design engineer must make in the initial specification of the materi-
als to be used for any particular application, recognizing that the
creep potential of the component must be countered by other design
modifications or parameters.

496
Steam Path Damage Induced by Material Degradation

THE CREEP MECHANISM


Creep deformation occurs in all components subject to stress,
and operating at any temperature. In practical applications, only
those components at an elevated temperature and subject to high
stress levels are of significance and interest to the turbine engineer.
(There are threshold values of temperature and stress that vary with
the material being considered.) This interest derives from the fact that
those components mechanical condition, performance, and
remaining life must be observed and possibly monitored.

The term elevated temperature is used without definition. Its


significance in the expectation and examination of creep and the
temperature at which significant creep can be expected to occur is
temperature that is about one half of the metal-melting temperature
on the absolute temperature scale (K = F + 460), i.e., If an alloy
steel has a known melting point of 1,120F, then significant creep
levels can be expected to occur at a temperature of:

0.5(1,120 + 460) = 790F

Note: The majority of turbine steels melt between 1,300F and


1,400F. When they are used in the high-temperature range, it is nor-
mal to consider that creep deformation could occur in those com-
ponents with an operating temperature at or above 750F, and their
design is considered for creep deformation. This provides a built-in
factor of safety.

Such components are normally designed to provide a minimum


creep life of 200,000 hours, which gives an operating life of about
30 years, at a load factor of 75%.

We know that relatively low-magnitude stresses are capable of


producing creep deformation when applied for a sufficient time at an
elevated temperature. Under these conditionselevated tempera-

497
Turbine Steam Path Maintenance and RepairVolume One

ture and low-magnitude stress applied continuouslycomponents


deform gradually. They begin at a high rate, which continues until
the lower secondary level of deformation occurs. At this secondary
condition, micro-cracks occur both at the surface and internally, and
can grow until a tertiary rate of deformation is entered, which con-
tinues at ever-increasing rates until final rupture occurs. The creep
rate is, in effect, the rate at which the component deformselongat-
ing, compressing, or even bending. It is a measure of the rate at
which the component shape changes in the direction resulting from
the application of the stress.

It is of interest to consider the internal mechanisms, ruptures, and


crystal adjustments that occur in the material under conditions of
load and temperature. It is these changes that contribute to, or cause
and promote this permanent plastic deformation.

Consider the schematic (Fig. 5.5.1), representing a micrograph


section through a piece of material. It is at a constant elevated tem-

Tensile Elastic
Load "L" extension
Original length

Tensile
Load "L"

Figure 5.5.1The schematic


Figure 5.5.1 of material
placed
The in tension
schematic with a small
of material elastic
placed in tension
extension.
with a small elastic extension

498
Steam Path Damage Induced by Material Degradation

perature and subjected to a tensile load L. This load is of a size that


neither rupture nor significant deformation occurs as it is applied
suddenly. There is shown to be a small elastic deformation from the
original length, but this is such a small amount that a major portion
would be recovered if the load were removed after only a short peri-
od of application (elastic extension). However, after an extended
period, further deformation will be notedthis deformation com-
prising any initial elastic and a more permanent plastic deformation.
This deformation can be the result of two separate forms of structur-
al change within the material occurring together, and are illustrated
in Figure 5.5.2. They are explained below.

The existence of impurities (inclusions) or micro-voids at the


grain boundaries

Figure 5.5.2The effect of extended service on material structure at an


elevated temperature, with the material subject to a tensile stress.

499
Turbine Steam Path Maintenance and RepairVolume One

Consider the boundary that is close to perpendicular to the


direction of the applied load L, between grains A and B. It indi-
cates there are a certain number of impurities or even small voids
present at this boundary.

As the load is applied, a tensile load develops across this bound-


ary, and this face is placed in almost direct tension. The voids and
impurity centerswhich exist at this surface and normally have less
resistance to applied loadwill tend to grow along the boundary,
perpendicular to the applied load. This growth will continue until
adjacent voids join, and in sufficient length can be considered to
form a micro-crack.

After a micro-crack forms, the remaining connected length of the


grain boundary or boundaries must take up a portion of the total load.
This causes an increase in the internal tensile stress across all remain-
ing load-bearing boundary surfaces. If the micro-crack grows for the
entire length of the grain boundarypossibly extending over several
gains of the materialthese grains are unloaded and the total load
must be borne by those grains that remain attached. It is normal to
expect that grain boundaries with voids and impurities will not all
unload at the same time. Therefore, while the external load is
unchanged the internal stress will increase, allowing further cracks to
initiate and grow causing a further increase in the initial deformation.

If the load is removed, the width of the void and impurity crack
will not decrease significantly. The deformation will not disappear.
This (plastic) deformation is permanent.

As more voids connect, the load and local stresses rise, promot-
ing a higher (faster) crack growth rate in the remaining connected
grains. This form of crack formation is inter-granular, and de-cohe-
sion of the voids will exist at the grain boundaries.

500
Steam Path Damage Induced by Material Degradation

Wedge cracks formed at grain boundary junctions, as shown


for the common junction of the boundaries of grains C,
D, and E

These wedge cracks extend along the grain boundaries, again


unloading the affected grains and placing the load on the remaining
portion of the material still connected.

Like those micro-cracks formed by void de-cohesion, this type of


crack is inter-granular. However, it should be expected that those
boundaries most closely inclined to the perpendicular (to the
applied load) will suffer the greatest stress effect, and it should be
expected that any micro-cracks that form will preferentially develop
along those boundaries such as those existing between grains C
and E in Figure 5.5.2.

As these various voids coalesce and greater loads are placed on


the remaining material, this remaining material between the micro-
cracks will undergo considerable plastic deformation before final
rupture occurs.

The most common cause for creep deformation (considered in


the previous sections) is when a tensile load is applied. However, it
is not uncommon for extension to occur under other forms of load,
e.g., compressive, shear, bending, or torsion. Under these types of
load, plastic deformation will occur if the conditions of stress and
temperature are such that they promote the creep phenomena.

Grain and boundary layer relative strength


When a load is applied to a piece of material, and conditions of
temperature and stress are such that creep deformation occursthe
deformation being caused by the formation of micro-cracks and the
deformation this allowsthere is a question of the relative strength
of the pure grains of the material, and the material contained within
the boundary layers.

501
Turbine Steam Path Maintenance and RepairVolume One

At lower temperatures, grain boundary material tends to be


stronger than the materials of the eutectic grains. Therefore, at lower
temperatures, ruptures occurring will tend to be trans-granular, while
at higher temperatures they will be inter-granular. The temperature at
which failures move from the trans-granular to the inter-granular
form is called the equi-cohesive transition temperature. This tem-
perature is dependent upon a number of variables, such as the stress
applied and the amount of time the component has been exposed to
the temperature at which it is operating.

Figure 5.5.3 shows a diagrammatic representation of the variation


of strength of the grains of pure material and the grain boundaries.

Strength
of grain
boundaries
Measured Material Strength

Strength
of material
grains

Equicohesive
transition temperature

Temperature

Figure 5.5.3The relative strength


Figure of the5.5.3
individual metal crystals and the
grain boundaries as astrength
The relative functionofofthe
temperature.
individual metal crystals and the
grain boundaries as a function of temperature.

Slip planes
Slip is a form of plastic deformation that occurs under the
action of shear loads, causing a permanent displacement in the
direction of the load. When slip occurs in a specimen subjected to
creep deformation, slip planes tend to be along the grain boundaries

502
Steam Path Damage Induced by Material Degradation

under the load intensity. Such slip would be visible by microscopic


examination or using scanning electron microscope methods.

If the material section shown in Figure 5.5.1 has a shear load


applied (Fig. 5.5.4), a plastic deformation will occur through some
angle (as shown), giving a total deflection . Again, this deforma-
tion occurs because of the presence of the two parameters of stress
and temperature previously discussed. Such deformation will be
plastic and therefore permanent.

Shear load "S"

Shear load "S"

Figure 5.5.4
Figure 5.5.4Material deformation when
Material
a shear deformation
stress is applied.when a shear

Creep progression
Figure 5.5.5 shows a schematic of creep damage growth. This
growth damage from individual voids or cavities connect to form
these cracks and continue to connect until the micro-cracks join,
forming major macro size ruptures that eventually produce complete
failure.

503
Turbine Steam Path Maintenance and RepairVolume One

Figure 5.5.5Schematic of the progression of creep damage from isolated cavities to


macro-cracking.

CREEP RATE
As previously discussed, the creep rate is not constant under any
applied load, even when the component operates at a relatively con-
stant temperature and levels of stress. Rather, there is an initial high
rate of primary deformation that slows through a secondary phase,
and is maintained relatively constant for considerable periods,
before entering a tertiary phase where deformation is again rapid to
failure.

It is interesting to consider the material characteristics that pro-


duce these changes in creep rate. Unfortunately, no precise expla-
nation can be found, but if we consider the crystallographic and
property changes occurring within a component subjected to the
conditions promoting (or inducing) creep, some explanation for
creep rate can be established in general terms.

504
Steam Path Damage Induced by Material Degradation

When load is applied to a component at elevated temperatures,


there are degrees of work hardeninga process by which the
hardness and mechanical strength of a material increases as a con-
sequence of plastic deformation. This phenomenon occurs below the
re-crystallization temperature and reduces the creep by increasing
the materials ability to oppose deformation. When temperatures are
below the re-crystallization range where turbine components oper-
ate, it is still possible for re-crystallizationa process by which large
grains are refined to form a number of the smaller crystalsto occur
slowly. In turbine steels this is a form of cold working that changes
the crystal structure at the localized temperature, and tends to
increase the creep rate.

A further material characteristic change that can influence the


creep rate is precipitate over-aging. This phenomenon produces an
increase in certain material mechanical properties in the direction
the load is applied. This increases the creep rate.

How these various phenomena interact in a material to modify


the creep rate depends upon time and operating environments.
During short time applications of load, work hardening is the pre-
dominant characteristic. This means the initial rate of creep is high,
and deformation rapid, in the direction the load is applied. However,
as the time-of-load application increases, those phenomena that
tend to lower creep rate begin to become effective. A balance is
reached between them and the work hardening effect to the extent
the secondary phase of creep rate is relatively constant. As time
increases, there will be an accumulation of damageformation of
surface and internal cracksand the creep rate will accelerate into
the tertiary phase, where the rate of material deformation and dam-
age accumulation increases to final rupture.

505
Turbine Steam Path Maintenance and RepairVolume One

Prediction of creep rates


The methodology available for predicting creep rates is not
exact. However, there is sufficient experimental data available that
certain empirical relationships have been formulated, sufficient that
a reasonable level of accuracy in rate prediction can be achieved.

There are three basic equations defining the instantaneous, pri-


mary, and secondary strains:
1/3 .
Instantaneous = o + C1 t +t
.
Primary = o + C2 1 - exp ( -C3t + t
.
Secondary = o + ln { 1 + C [ 1 - exp ( - C4 t)]}
C4 5

where:

o is the instantaneous strain


t is time in hours
and
C1, C2, C3, C4 and C5 are empirical constants dependent upon the
mechanical and chemical properties of the material

In the tertiary stage of creep the strain is best defined by equations:

where:

exp [ C6 ( t - t )]
t 3
1

t C7
( 1 - D)

t is the time in hours


t3 is the time in hours to enter the tertiary range
D is a constant that varies with time

Note: The constants defining these deformation rates are materi-


al dependent, which can be defined.

506
Steam Path Damage Induced by Material Degradation

CREEP IN STEAM PATH


COMPONENTS
Since many components of the turbine steam path are subjected
to continuous loads and elevated temperatures (elevated as defined
earlier), then it should be expected that some level of creep defor-
mation will occur during their operating life. This section will con-
sider those components susceptible to this form of damage.

Figure 5.7.1A control stage blade root having suffered creep


damage.

507
Turbine Steam Path Maintenance and RepairVolume One

Rotating blades
The rotating blades are subject to continuous loads during oper-
ation. In the early stages of the high and reheat sections of the unit,
these loads are applied at temperatures sufficiently high that creep
deformation can and will occur. The extent of this deformation and
the need to replace blades as a consequence is dependent upon a
number of factors, e.g., the magnitude of the load, and the time for
which the blades have been in operation. Figure 5.7.1 shows a blade
from the control stage of a high-pressure section with initial steam
temperature of 1,100F, and which has been in service for 160,000
hours. This blade has lasted remarkably well, but under the influence
of stress and temperature over an extended time, has ultimately crept
until the load bearing surfaces on the root ligaments have started to
move outward past the transfer surfaces on the rotor. This action has
sheared the side grips. Shown as Figure 5.7.2 is the form of the creep
deformation of a similar blade root, indicating the initial and final
operating condition, with the lift due to creep indicated.

Another factor that can induce damage in any component, but


especially on rotating blades, is failure to hold certain design speci-
fied dimensional tolerances. Figure 5.7.3(a) shows the closing blade
from the first rotating blade row of a reheat section, with a reheat
temperature of 1,000F. In this unit a false start was made in pro-
ducing the root pinhole. The consequence of this is shown in Figure
5.7.3(b)It did not have sufficient surface to transfer its load.
Therefore, under the influence of centrifugal loading, the pin distort-
ed or crept, allowing the blade to migrate radially outward.

In general, high-temperature rotating blades are manufactured


from an AISI 410 or 403 martensitic type stainless materials. There are,
however, some high-temperature/high pressure applications where
either an AISI 422 or an austenitic type material can be employed. A
concern with austenitic materials is they have a different coefficient of
expansion. It is necessary to account for this difference during the
design phase to ensure that under thermal transient conditions the unit

508
Steam Path Damage Induced by Material Degradation

Figure 5.7.2Creep deformation in a straddle type


root of the form shown in figure 5.7.1

is able to accommodate these differences, and their possible effect on


unequal load sharing between the load bearing surfaces.

Note: When replacing rotating blades for a stage in the creep


range, it is preferable to determine any creep deformation that has
occurred, as producing blades to the original dimensions may not
allow equal load sharing between multiple load transfer surfaces.

509
Turbine Steam Path Maintenance and RepairVolume One

Rotors
Rotorslike other components subject to high temperature and
high pressureare subject to creep deformation. This damage
occurs most often in regions of the rotor where stresses are high.
These are most often regions where load is transferred from the blade
root to the rotor rim. Figure 5.7.1 shows a wheel rim where the blade
root has deformed and blades detached. This rim has also deformed
due to creep.

Figure 5.7.3(a)Root creep see figure 5.7.3(b).

510
Steam Path Damage Induced by Material Degradation

Figure 5.7.3(b)Showing the false start on the pin


hole at manufacture, permitting blade root creep.

High-temperature rotors are of a complex form and the highest


temperature and stresses may not occur at the same location.
However, there is normally sufficient coincidence between these
two contributing factors that creep has a high probability of occur-
ring, if only at a relatively low rate.

The rotor can also suffer damage as a consequence of combined


creep fatigue. See HCF Failure Surface Appearance.

Diaphragms
Diaphragms are produced in three partsan outer ring, a blade
annulus consisting of a number of vane elements, and an inner web.
Each of these three components of the diaphragm is subjected to the
pressure differential of the stationary blade row. The outer ring is sup-

511
Turbine Steam Path Maintenance and RepairVolume One

ported against the locating slot in the casing, and does not distort
under the effect of pressure. However, both the vanesproduced
from a martensitic steel of the same general mechanical and chemi-
cal properties as the rotating bladesand the inner web are subject
to pressure differential, and will eventually take up an elastic and, if
at sufficiently high temperatures, plastic or creep deformation.

Y Y

Figure 5.7.4Diaphragm deflection,


Figurebeing
5.7.4a maximum at the inner diame-
ter on the horizontal
Diaphragmjoint.
deflection, being a maximum at the inner
diameter on the horizontal joint.

As steam is initially admitted to the unit, stresses are set up in the


diaphragm. These initial stresses induced in the various components
of the diaphragm are of little significance in themselves, their mag-
nitude being in the order of 5,000-10,000 psi. However, the axial
loads, while not high, result in a down-stream deflection of the
diaphragm that is a maximum in the inner web inner diameter adja-
cent to the rotor. Figure 5.7.4 shows the deflection of a diaphragm at
the inner diameter. The maximum stresses that occur in a diaphragm
due to the pressure differential it experiences, occur in the element
along the axis XX, which is at 90 to the YY axis or horizontal

512
Steam Path Damage Induced by Material Degradation

joint. It can also be seen from this diagram that the maximum deflec-
tion does not occur on the axis of maximum stress, but rather on the
horizontal joint which, during operation, receives less support from
adjacent material than does the vertical center line XX.

Outer
Ring Casing

Stationary
Vanes
Rotor
Outline
Initial cold
Inner clearance
Rotor Web "Cli"
Outline Final hot
running
clearance
Initial Cold Setting "Clr"
Elastic Deflection.
Creep at time "T1"
Creep at time "T2"

Figure 5.7.5The stages


Figure of deflection
5.7.5 of a diaphragm
with time
Theexposure.
stages of deflection of a diaphragm
with time exposure.

The outer rings and webs in the high-temperature regions are pro-
duced from forged, high-strength alloy steel. The major concern with
the manufacture of diaphragms is the ability of these components to
resist for extended periods at high temperatures, the axial pressures
developed across them. The outer ring is located in the casing to form
a steam-tight joint that will not deflect to any significant degree.
However, both the vanes and inner web will deflect downstream
under the influence of the axial pressure gradient. It is therefore nec-
essary to consider the effect of time on operating deflection and the
influence this will have on axial clearances within the steam path.

513
Turbine Steam Path Maintenance and RepairVolume One

Diaphragms have, compared to the rotating components, rela-


tively low levels of stress induced in them. These components are
unlikely to fail in a catastrophic manner. However, when steam is
admitted to the unit, there is a pressure differential stress induced in
the vanes and inner web that will cause it to deflect downstream.
When a unit goes into service, each diaphragm will experience an
initial elastic deflection followed by a time and temperature depend-
ent plastic deflection. Figure 5.7.5 shows a cross section of a
diaphragm in its operating position relative to the rotor. It can be
seen that:

there is an initial cold clearance between the inner web, at


its inner diameter, and the rotor Cli, when the rotor is in the
stationary condition

when the unit starts and steam is admitted, there will be an


initial elastic deflection, under the influence of the stationary
row pressure differential, consuming a portion of this clear-
ance reducing it to Clo

after a time T there will have been creep deformation,


which is an additive to the elastic deformation. After this time
period, the creep deflection will have increased, reducing
the clearance to Clr

As operation continues, the initial hot running clearance is


reduced by a deformation of the creep deformation, and a condition
could eventually exist that a rub will occur between the
diaphragm inner web and the rotor. This rub is normally at the
diaphragm inner diameter, or at any location where tight running
axial clearances might exist. Such a rub will generate a considerable
amount of frictional heat. These heated regions are immediately
quenched by the surrounding steam and introduce a hardened zone
on both the diaphragm and rotor. While the condition of the
diaphragm may not be immediately serious, the rotor will be burnt

514
Steam Path Damage Induced by Material Degradation

to the extent it will become particularly hard, and brittle, and should
not be operated without corrective actions being taken.

This rub can often occur during a start up or shut down situ-
ation, when the rotor could be subjected to short and long con-
ditions (see chapter 2) modifying the normal operating clearances.

Note: When the unit has stopped, and the steam shut off, the
elastic deflection will be reduced to zero. However, the plastic
deflection is permanent, and will not be recovered. It is recom-
mended to monitor critical clearances as a part of the normal main-
tenance record in stages operating in the temperature region, where
creep could be present.

In an effort to increase resistance to creep deflection, and to


reduce the extent of initial plastic deformation, some manufacturers
produce their steam path stationary vanes from different profiles. The
vanes used in those stages having a predicted high creep deforma-
tion are evaluated for the effects of pressure, temperature, and steam
momentum loads. If these loads are excessive, it is common to use
extended axial width vanes as shown in Figure 5.7.6. Such vanes are
used on a portion of the total in each stage, and extend the axial
depth of the steam path from W by an amount E, to Wu.

Figure 5.7.6Extended section vanes, for axial strength, within a


steam path.

515
Turbine Steam Path Maintenance and RepairVolume One

The extended portion of these vanes is selected to preserve, to


the greatest extent possible, the aerodynamic form of the expansion
passage, and has little or no effect on the flow distribution of the
steam. This vane extension has little or no influence on the perform-
ance of the stage, because the velocity of the steam entering from the
previous rotating row is low and practically at 90 to the tangential
direction. However, if the steam enters at some angle , as shown
in Figure 5.7.7, this could change the amount of steam entering each
expansion passage, and cause minor changes to the stage flow char-
acteristics, having an effect on the swallowing capacity of the indi-
vidual passages.

Figure 5.7.7The effect of axial stiffeners on steam flow distribution in the


event steam does not enter axially.

Note: This change in the quantity of steam entering the individ-


ual nozzle passages modified their swallowing capacity. As a conse-
quence, there will be a difference in the steam discharge velocity,
pressure, and direction from adjacent passages in the row.

Diaphragms can also creep in the radial plane and affect their
horizontal joint. This effect is seen in Figure 5.7.8 where both diam-
eter increases and decrease are shown. This effect, and corrective
actions are discussed in chapter 7. In Figure 5.7.9 is shown the

516
Steam Path Damage Induced by Material Degradation

clamped halves of a diaphragm, before correction, where the halves


have moved in, opening a gap at the horizontal joint.

Original Design
Position.

Steam
Path

Original Diameter D.
dR = dD/2

Original Design
Position.

Steam
Path

Original Diameter D.
dR = dD/2

Figure 5.7.8Diaphragm radial creep


Figure deformation. In (a) the
5.7.8
diaphragm
Diaphragm has crept
radial inwards
creep reducing
deformation. In the diameter
(a) the by anhas
diaphragm amount
crept-
dR,inwards
and in (b) the diaphragm
reducing has opened
the diameter by a similar
by an amount "-dR",amount +dR.
and in (b)
th di h h db i il t " dR"

Note: When considering the axial or radial distortion of


diaphragms it must be considered that creep is not the only mecha-
nism introducing this condition. It is also possible that residual
stresses locked into the component due to a welding or other process
could cause this condition to occur. It is also necessary to recognize
while this condition can be corrected, in cases of creep where dis-
tortion is removed, voids that exist between the crystals of the mate-
rial will not be removed and deflection will occur again at a rela-
tively fast rate.

517
Turbine Steam Path Maintenance and RepairVolume One

Figure 5.7.9The diaphragm horizontal joint gap caused by the halves moving in due to
creep deflection.

Casings and blade carriers


Turbine casings and blade carriers are subject to high levels of
stress and high temperatures. Under these conditions there can be a
degree of creep deformation causing the circular form of the casing
to be modified to an elliptical form, with the horizontal joint either
increasing or reducing. The possible consequences of this deforma-
tion are dependent on details of the design. In certain circumstances
such deformation can cause serious damage and even introduce dif-
ficulties in removing the studs and lifting the top half casing or blade
carriers, because of binding between the casing and horizontal joint
studs. In other situations the difficulties involved are minor, will not
influence steam path alignment, and can therefore be tolerated.

Note: Cylinders that are hotter on the inside surface than the out-
sides tend to move-in at the horizontal joint. This can make the
removal of horizontal joint bolts difficult, and has been known to
make unit disassembly difficult, even resulting in the destruction of
a large number of studs. In the worst cases, even after the top half

518
Steam Path Damage Induced by Material Degradation

casing has been removed, the lower half will move in and grab the
rotor, making its removal difficult without causing damage to the
rotating blades and shaft end seals.

Figure 5.7.10The inner surface of a casing where creep deformation has moved it in
to the extent heavy rubs with the rotating blade integral seals have grooved the casing.

Figure 5.7.10 shows the horizontal joint of a casing that has


moved in at the horizontal joint, producing heavy rubs between
the casing inner surface and the integral seals produced on the blade
tips. In this design the stationary blade rows are carried in the inner
casing, and so any distortion it suffers will cause a misalignment
between the stationary and rotating blade row, reducing and ulti-
mately destroying the lap. This type of stationary blade row defor-
mation can also produce a tangential variation of the steam force,
possibly introducing stimuli into the rotating blades, causing some
form of vibration.

Note: As in the case of diaphragms, the distortion present in cas-


ings can be due to phenomena other than creep, and the comments

519
Turbine Steam Path Maintenance and RepairVolume One

regarding their correction and future deformation are also valid for
casings.

Threaded components
Many studs and bolts are always in a state of high tension, and
operate at high temperatures. If the components screwed ends are
contained within a thick horizontal flange it is unlikely they will
extend by any significant degree. Similarly, a thickheaded nut will
not allow significant deformation. However, the shank can and will
extend with time. This deformation should be monitored as the life
of these components should be considered finite and should be
changed whenever there is an indication that their life has expired.

HIGH-CYCLE FATIGUE
High-cycle fatigue is a mechanism capable of causing mechani-
cal rupture in a component as a consequence of the application of
repeated cyclic loads. These are loads whose maximum values can
be considerably lower than those required to cause failure by a non-
alternating application, when applied without shock, and in the
same direction to the same component.

This mechanism of failure is perhaps the most common observed


in failed components of the steam turbine. The high frequency of this
type of failure is a consequence of the many sources of cyclic load-
ing generated within the steam flow. These cyclic loads, while of low
magnitude, are of high frequency in a component that is rotating at
high speed, and therefore can accumulate many stress cycles within
a relatively short period. In addition, there can be present many
regions of stress concentration as a result of damage caused by some
other initiating mechanism. Components normally able to sustain

520
Steam Path Damage Induced by Material Degradation

the initial form of damage and be unaffected would fail as a conse-


quence of the application of a low-magnitude alternating load.

The design process establishes the dimensional and material


requirements for each component of the steam path, and selects
them so stress levels (direct and alternating) induced by operation
will not exceed values that can be tolerated by the components. It is
often relatively minor damage that will initiate damage, and then
high cyclic loading will drive the components to failure.

High-cycle fatigue is most destructive in regions where high


direct stress levels exist. The most common location to discover high
stress levels is where there is stress concentration. There are three
basic causes of stress concentration:

Section changes. Design considerations often require section or


form changes. These changes can introduce regions where small fillet
radii or other form changes are required. This causes load to be
applied in a non-uniform manner, which increase stress levels locally.

Material deficiencies. As materials are produced for turbine


application, they are carefully examined to ensure faults, (e.g., inclu-
sions) do not exist in regions where they represent a possible com-
promise to the material quality. However, it is possible for such faults
to go undetected. It is also possible for faults to be introduced par-
ticularly during any manufacturing process involving the use of heat.
These faults (e.g., hard spots) can go undetected despite the best
efforts of the manufacturer to eliminate them.

Machining or assembly marks. There are many manufacturing


and assembly processes where surface marks, scratches, and gouges
can be introduced. Some of these may even be hidden at completion
of the process, making them more difficult or impossible to detect.
These can act as stress concentration centers, capable of causing
damage.

521
Turbine Steam Path Maintenance and RepairVolume One

THE HIGH-CYCLE PHENOMENA


The alternating stresses applied to a component vary between a
maximum and minimum, and vary from a mean value. Therefore, if
the mean stress is denoted by m, and the alternating stress has a
range of 2a, then the maximum stress to which the component is
subjected is m+a, and the minimum m-a. This varying
stress effect is shown in Figure 5.9.1. In fact in practice there are
three forms (or stress ranges) in which this stress is applied:

Figure 5.9.1An alternating load of 2a applied to a constant load of m.

The alternating stress changes from a maximum to minimum


around a zero mean stress. This is shown in Figure 5.9.2(a).
Therefore, the direction of the resultant stress varies during
each cycle

The alternating stress varies about a mean stress. Its lower in


magnitude than the mean stress and the total stress is at all
times in the same direction. This is shown as Figure 5.9.2(b)

The alternating stress varies from a maximum to a minimum,


the minimum being equal to zero. This is shown in Figure
5.9.2(c)

522
Steam Path Damage Induced by Material Degradation

Figure 5.9.2The three forms of alternating load application.

In defining these three formsdirection and magnitude of alter-


nating stress level and its variation, as depicted in Figure 5.9.2it
has been assumed the stress is applied in a sinusoidal form and max-
imum and minimum values of the applied alternating stress are
always of the same magnitude. These basic assumptions are not nec-
essarily true. Within any steam path, there can be considerable vari-
ation from them.

523
Turbine Steam Path Maintenance and RepairVolume One

The manner and frequency in which the varying loads are


applied is dependent upon a number of factors related to the steam
path geometry and its compliance with the standards set by the
designer. Important factors in establishing the magnitude of the
steam force include but are not limited to the following issues:

The strength of the wake at discharge from between two sta-


tionary blade vanes. This can be dependent upon compli-
ance of the blade vane

The inlet conformance of the individual passages in the sta-


tionary row and the differences in their swallowing capacity

The gauging of the individual throats within the steam path,


which will modify the steam discharge angle

Any inconsistencies that might exist in steam path form at the


horizontal joint of the stationary blade row

The magnitude of the impulse will also be affected by the quan-


tity of steam flowing at any time. This is dependent upon the load the
turbine is required to generate, which obviously varies with system
demand on that unit.

It has been assumed in Figure 5.9.2 that the frequency of appli-


cation of the stress remains unchanged. Within the turbine steam
path of a unit designed for constant frequency power generating this
is a valid assumption. However, in other turbines designed to oper-
ate at varying speeds, this cannot be assumed. Because the accumu-
lation of damage is frequency sensitive, such units can accumulate
different amounts of damage at different times. The most damaging
situation to arise in the steam turbine is when the frequency at which
the impulses developed within the steam path is coincident with the
natural frequency of the components these impulses are interacting
with. Under these circumstances the impacted component can
vibrate to failure with a relatively short period of time.

524
Steam Path Damage Induced by Material Degradation

From these considerations it is clear that an analysis of the steam


path alternating stress levels can be complex. A normal procedure in
the design phase is to select and dimension the various components
to fall within certain limitations. Then, should damage be sustained in
operation from some form of vibratory loading, at that time a detailed
analysis can be made to identify the circumstances of failure.

ROTATING BLADE
VIBRATORY STRESSES
The blades of a turbine row are usually the most susceptible to
high-cycle fatigue, and have induced in them stresses caused by the
vibratory stimuli developed within the flowing steam during opera-
tion. The magnitude of these stresses vary from stage to stage, and are
influenced by the frequency and magnitude of the vibratory load
sensed. The physical characteristics of the bladesize, vane form,
and the material from which they are produced (physical properties
and damping characteristics)will also influence the total effect of
the stimuli on the blade. Various sources of vibratory stimuli in the
steam path are given in Table 5.10.1.

525
Turbine Steam Path Maintenance and RepairVolume One

Section Harmonic Typical Sources

High High per rev Nozzle tolerance limits


Pressure (40x) Upstream wake
Section degeneration
Structural turbulence

Intermediate Nozzle passing Nozzle wakes


Pressure frequency (NPF)
Section 2 x NPF Diaphragm harmonics
3 x NPF Diaphragm harmonics

Low One per rev Relative displacement


Pressure nozzles to blades
Section 2 per rev Diaphragm joints
Multiple/rev Structural supports
in flow path
Medium/rev Diaphragm harmonics
Aeroelastic disturbances
High/rev Nozzle turbulence
harmonics
Upstream wake
degeneration
Structural turbulence

Table 5.10.1Steam Path Exciting Forces

These various modes of vibration can cause the blade to vibrate


in the axial, tangential, and torsion modes. There are various har-
monics of these modes, and the magnitude of any induced stress is
a maximum when the frequency of the impressed force, or stimulus,
is in phase with one of the natural frequencies of the blade itself. This
coincidence causes an increase in the vibration amplitude, and the
strain and stress this induces.

The forces acting on the blade and contributing to vibratory load-


ing are the steam bending forces developed as the result of the steam
flow across the vane. Many of these loads are predictable and the
design process avoids introducing these into the blading at a level

526
Steam Path Damage Induced by Material Degradation

where they can cause failure or damage. While the magnitude of


these steam forces may be low, if their frequency of application is
close to a natural blade frequency, they can soon induce a failure
into the rotating blade elements.

Nozzle impulse effects


A common cause of damage or failure in the turbine rotating
blade system is the effects of nozzle impulses. A nozzle impulse
results from fluid flow distortions. Distortion results from uneven
flow patterns between the stationary blade vanes. As the rotating
blade vane traverses the nozzle pitch, it is subjected to a variation of
steam pressure impulses, and therefore the forces developed across
its tangential and axial length. A typical impulse/force diagram for
three nozzle pitches (four vanes) is shown as Figure 5.10.1, where
the steam force F on the blade at any instant is varying dependent
upon its position in the tangential direction across the blade outlet.

Figure 5.10.1The variation of steam force f across the pitch P of the fixed blade
row.

527
Turbine Steam Path Maintenance and RepairVolume One

Also, shown in Figure 5.10.1 is the effect of a viscous trailing


edge wake. The magnitude of the force produced by this wake effect
is dependent upon several factors, including the flow passage shape,
and the extent to which the boundary layer separates from the noz-
zle trailing edge. Shown in the lower portion of the diagram is a
schematic of the steam force F at discharge from the blade row,
drawn above some arbitrary position. The actual steam pressure
decreases across the nozzle passage, and at the discharge edge,
where wakes (We) form, there is a small reduction in pressure
due to this effect. This steam force can be seen to be acting on the
vane. Because it is a varying force, the entire blade is subject to an
alternating force, which induces a cyclic stress of varying magnitude
across each nozzle passage.

For rotating blades, this nozzle impulse effect is possibly the


most common cause of failure. For most designs the normal magni-
tude of the combined tensile and bending stress are not high enough
to cause failure. However, the effects of this varying nozzle force can
be magnified to produce dangerous levels of alternating stress.

Nozzle passing effect (frequency)


Blade resonance occurs when the natural frequency of the blade
is at or near coincidence with the frequency of the impressed forces
developed by the pressure variation across the nozzle discharge
shown in Figure 5.10.1.

The magnitude of the steam force produced by the nozzle dis-


charge pressure variation is relatively small, and of little conse-
quence when compared to the rigidity of the blade. For this reason,
if the natural frequency of the blade is not at or near coincident with
the nozzle passing frequency, the blade dynamic stresses will not be
influenced to a considerable extent.

528
Steam Path Damage Induced by Material Degradation

To quantitatively establish this effect, an amplification factor to


be applied to the steam bending force can be determined.

Consider the blade pair from a rotating row shown as Figure


5.10.2. It is subject to nozzle impulses. Consider each to act as a can-
tilever under the influence of these impressed forces. Also assume the
blade has internal damping due to its molecular structure.

F. Cos t

Figure
Figure 5.10.2The steam5.10.2
force developed on
the rotating
Theblade
steamrow.
force developed on
the rotating blade row.

The equation of motion of the blade is given by:

2
d y dy
M 2
+ Fd + Ky = F cos t
dt dt

where:

K = Blade stiffness
M = W/g
Fd = Internal damping force
(Numerically = Magnitude of damping force
when the velocity is unity)
Y= Displacement of the blade from equilibrium in time t

529
Turbine Steam Path Maintenance and RepairVolume One

Velocity coefficient is v, and v = Fd. dy/dt. This can also be


written in the form:

2
d y dy
+a + by = c
2 dt
dt

The complete solution of this equation consists of a decaying


time-transient portion (the complementary function) and a sustained
portion (the particular integral) as follows:

y = ys + yt

In this equation, the first term represents the transient vibration,


which can be assumed to be completely dampened by molecular
friction. The second term represents the forced vibration and is main-
tained by the periodic nozzle force. Once a steady state is attained,
this vibration is represented by:

c
y = . cos t -
2 2 2
b- + a
Fd K F
Where "a" = , "b" = , and "c" =
M M M

y has a maximum value when cos (t - 1) = 0, etc. c is neg-


ative because initially the response y lags the driving force, c =
F/M. Then:

c c/b
y max = =
2 2 2 2
b- + a 2 2 2
a
1- +
b 2
b
Now, c/ b = F/K =

530
Steam Path Damage Induced by Material Degradation

Now, c/b = F/K = (where is the deflection produced by the


static force equivalent.

Fn is the nozzle passing frequency


Fr is the blade natural frequency

where:

Ad is the dynamic amplification factor

1
Ad =
2
2 2 2
1-
+
a
b 2
b

Which reduces to:


1
Ad =
2
2
Fn
1- + Cd
Fr
2
a
=
2 f
R

where: Cd is the damping coefficient

Figure 5.10.2 shows the steam flow from stationary blades into a
rotating blade row. Here the force onto the blades is shown as F.cos
t, produced at a frequency of Fn cps. Depending upon the nat-
ural frequency of the rotating blades there can be considerable
amplification in the magnitude of vibration. Figure 5.10.3 shows the
amplification factor Ad shown in the equation immediately above.
This is shown as a function of the ratio blade natural frequency at
operating speed Fr to the nozzle passing frequency Fn. These
curves show the influence of various damping coefficients Cd on
the amplification factor, and therefore, the dynamic stresses devel-
oped in the blade.

531
Turbine Steam Path Maintenance and RepairVolume One

Figure 5.10.3The dynamic amplification factor Ad, as a function of the ratio of


nozzle passing frequency to blade natural frequency for a range of damping
coefficients Cd.

Combined stress
In evaluating total stresses on a blade it is necessary to be able
to determine the effects of the combined static (centrifugal), and
vibratory (steam bending) stresses. A convenient method for doing
this is to use the Gerbers Parabola or Gerbers Diagram. Gerber
made an analysis of the fatigue results of various metals, and deter-
mined the effects of the combined stresses that could be represented
by Equation 5.10.6. This expression is shown on curve Figure
5.10.4(a), with the limiting and actual stresses as shown.

where:

532
Steam Path Damage Induced by Material Degradation

2
f
' = Fv 1 - 5.10.6
T

= The alternating (limiting steam bending) stress


Fv = The limiting range of alternating stress, = T/2
f = The mean direct stress = fc + fb/2
T = Tensile ultimate strength of the material
Alternating Stress

Direct Stress
The alternating stress.
The alternating stress limit.
The mean direct stress.
The ultimate tensile stress.
Alternating Stress

Direct Stress

Figure 5.10.4In (a) is shown Figure 5.10.4Parabola and in (b) the


the Gerbers
In (a) is shown
Modified Goodman the "Gerbers Parabola" and in (b) the
Diagram.

533
Turbine Steam Path Maintenance and RepairVolume One

This expression indicates that a parabolic relationship exists


between the combined direct and alternating stress and the
limiting stresses f and Lv that the blade can tolerate. It was sub-
sequently found that certain materials, including nickel and chromi-
um alloy steels, do not accord with this relationship, but rather enjoy
a linear relationship as defined by the modified Gerber or Goodman
equation 5.10.7. This Goodman straight-line curve is shown on
Figure 5.10.4(b).

' = Fv 1 - f 5.10.7
T

This expression is termed the modified Goodman linear rela-


tionship.

Both the Gerber and Goodman expressions can be represented


as shown in Figure 5.10.4. This diagram allows a representation to
be applied to the known stresses on the blade, and can be adjusted
to account for the dynamic amplification factor Ad of the alternat-
ing stresses, as determined from the frequency of the loads applied
to the stage. There are difficulties associated with determining exact-
ly all of the vibratory loads applied to a blade row, but predictive
methods exist. These can also be left to experimental determination
for a blade family profile, and the elements position within the unit.

To determine the factor of safety for the combined stresses, the


ratio OE/OF could be used as an indication of the blade acceptabil-
ity, based on either the Gerber or modified Goodman relationship.

Example 5.10.1
An alloy steel blade for operation in a 3,600-rpm unit has a pre-
dicted natural frequency of 5,360 cps when corrected for the stage
operating temperature. These blades are to be used in a row for

534
Steam Path Damage Induced by Material Degradation

which the designer has a choice of stationary blades elements to give


the correct discharge area and angle.

The available nozzle partition selection will be selected from:

Zn = 96 vanes - A
Zn = 84 vanes - B
or
Zn = 80 vanes - C

The predicted direct stress f is 13,350 psi; the limiting steam


bending stress (without amplification) is 4,640 psi. At stage
temperature the UTS (T) is 87,250 psi and the limiting fatigue stress
F is 47,750 psi. The blade material has a damping coefficient Cd
of 0.018. Its a matter of which stationary blade arrangement
appears the most satisfactory.

Solution:
Nozzle passing frequency Fn with 96 partitions
= 60 x 96 = 5,760 cps - A
Nozzle passing frequency Fn with 84 partitions
= 60 x 84 = 5,040 cps - B
Nozzle passing frequency Fn with 80 partitions
= 60 x 80 = 4,800 cps - C

The dynamic amplification factor, (applying Eq. 5.10.5) is:

For 96 partitions:(Fn/Fr)= 1.0746, therefore Ad = 4.88 - A


For 84 partitions:(Fn/Fr)= 0.9403, therefore Ad = 5.64 - B
For 80 partitions:(Fn/Fr)= 0.8955, therefore Ad = 2.15 - C

For each possible stationary vane application the direct stress is


22,350 + (Ad x ")/2

535
Turbine Steam Path Maintenance and RepairVolume One

For 96 partitions: 13,350 + 4.88 x 4640 = 24,672 psi - A


For 84 partitions: 13,350 + 5.64 x 4640 = 26,435 psi - B
For 80 partitions: 13,350 + 2.15 x 4640 = 18,338 psi - C

For each possible stationary vane application the alternating


stress is Ad."

For 96 partitions: = 4.88 x 4,640 = 22,643 psi Position - A


For 84 partitions: = 5.64 x 4,640 = 26,170 psi Position - B
For 80 partitions: = 2.15 x 4,640 = 9,976 psi Position - C

It can be seen from the Goodman Diagram, Figure 5.10.5 plot-


ted for the three possible nozzle vane arrangements. The amplifica-
tion factor Ad will dominate the choice for this stage. B with 84
stationary vanes is close to the Goodman limiting line, while A
with 96 vanes, is just a little safer. Again, this represents a marginal
design with a poor safety factor. However, C, with 80 vanes, is the
most promising with the coarser pitched nozzle partitions or station-
ary blade elements. This alternative C is therefore a safe design,
and represents an acceptable choice.

50
F

40
Alternating Stress

30
a
b
20

c 10

0
0 10 20 30 40 50 60 70 80 90
T
c ab Direct Stress

Figure 5.10.5The calculated direct and 5.10.5


Figure alternating stresses of the stage cal-
culated in example
The calculated 12.11.4
direct andshown on thestresses
alternating Goodman Diagram.
of the stage calculated in ex-
ample 12.11.4 shown on the Goodman Diagram.

536
Steam Path Damage Induced by Material Degradation

The control stage on a unit can be subjected to high dynamic


loading as a consequence of partial admission. These dynamic loads
result from the rotating blades moving into and out of arcs of admis-
sion and then into dead bands where there are no steam forces.
Where the nozzle passing effect represented a condition of a rip-
ple of steam force at admission to the rotating blade row, the effect
of partial admission is total removal and reapplication of the steam
force. This effect places heavy loading on the control stage, and must
be designed to withstand this condition, which is capable of causing
forced vibration irrespective of the natural frequency of the blade.

Partial admission effect


A simplistic diagram of the steam force, or loading diagram for
the single revolution of a blade, is shown in Figure 5.10.6(a). This fig-
ure shows the transient loading applied to the blade and the result-
ing deflection due to the steam jet effect.

Dead
One complete rotation of the moving blade row band
Steam
force

(d) (a) (b) (c) 2y (d) (a)

Figure 5.10.6(a)A simplistic diagram


Figure of5.10.6
the steam
(a) force diagram for one revolu-
tion 2 of the moving
A simplistic diagramblade,
of theshowing active
steam force arcsforand
diagram onedead bands.
revolution '2 ' of
the moving blade, showing 'active arcs' and 'dead bands'.

537
Turbine Steam Path Maintenance and RepairVolume One

In this figure, steam acts upon the rotating blade as it enters an


active arc at a. Here the steam jet deflects the vane, introducing a
stress in the vane proportional to the deflection. At the instant the
force is applied (position a) the kinetic energy of the steam jet
causes the blade to deflect, thus storing strain energy. This deflection
of the blade continues until the stored strain energy equals the kinet-
ic energy of the steam jet. At this condition the blade will have
deflected by an amount y. This kinetic energy will then dissipate
itself in the blade by deflecting it by a further amount y. At com-
pletion of this energy conversion, the total deflection of the blade
will be +2y, at which condition the stress due to this deflection
2y is equal to m. The blade bending stress will be twice that at
the equilibrium position. (This explanation assumes there is no inter-
nal molecular damping from the blade material.)

There will also be a small magnitude stress +/-a due to the


nozzle impulse effect.

If the blade being considered then rotates and moves out of the
active or admission arc at b, with the blade at the lower extremity
of its motion, the blade will then lose deflection, convert its stored
energy into kinetic, and deflect back to a position equal to -2y. The
blade will therefore vibrate at twice its normal amplitude. Under
such conditions, the stress range at the inlet and discharge edges will
be four times the static value.

If the blade then enters a second active admission arc c-d, with
its motion in the same direction as the steam force, the amplitude
will be further increased. This chain of events can continue until the
blade is ultimately destroyed. It is a requirement of design that the
blade is sized so its deflection will not produce stresses capable of
destroying it.

Actual stress levels can now be determined with considerable


degree of accuracy by finite element methods.

538
Steam Path Damage Induced by Material Degradation

The actual steam force diagram


The theoretical steam force diagram (Fig. 5.10.1) represents
forces as applied in a simple analysis. In fact, there are factors that
modify the form of this diagram.

First it is not possible for the force to be applied instantaneously,


and a finite time dti is required at inlet for the blade to sense fully
the axial effect of change in the steam pressure. Similarly, the force
cannot be removed from the blades instantaneously, and requires a
time dto to be removed completely. Therefore, the axial steam
force at positions a, b, c, and d [Fig. 5.10.6 (b)] will modify
in form.

Dead
One complete rotation of the moving blade row band
= 2 Radians
Steam
force

(a) (b) (c) (d) (a)


dT1 dT2

Figure 5.10.6(b)A modificationFigureof the simple


5.10.6steam
(b) force diagram to reflect the
effect of Aentry into and
modifcation exit
of the fromsteam
simple the fixed
forceblade rowtoactive
diagram reflectarcs.
the effect
of 'entry into' and 'exit from' the fixed blade row active arcs.

A second factor that modifies the force diagram at entry to an


active arc is the effect of the steam flow over only one surface of a
blade passage. This is termed the dip effect. Under the circum-
stances shown in Figure 5.10.7, the steam begins to enter the pas-
sage between the vanes. Before the blade senses the application of
load, there is a venturi effect on the vane suction surface that will
pull or deflect the vane towards the stationary blade row, tending to
unload it. This effect is shown as a negative force -dF1, which
exists just prior to the application of full load to the blade.

539
Turbine Steam Path Maintenance and RepairVolume One

Steam deflected
around the vane
inlet edge

Inactive portion
of the inlet arc.

Blade in the
dead band

Blade entering
the active arc
Steam -dF1
force

Figure 5.10.7The effects Figure 5.10.7


of entry to an active arc from a
The effects of 'entry to'
dead band, and the resultant forcesan active arcsuction
on the from a dead
face of the
band,
rotating bladeand the resultant forces on the suction face of
vane.
the rotating blade vane .

Inactive portion
of the inlet arc.

Blade deflected
+dF2 at entry to the
dead band

Steam Blade in the


force active band

-dF3

Figure 5.10.8The effect Figure 5.10.8


of enter to an inactive arc from the active,
andThe
the effect
effectsofofenter to an inactive
the steam arc from
load applied therotating
to the active, and the
blades.
effects of the steam load applied to the rotating blades.

540
Steam Path Damage Induced by Material Degradation

Similarly, as the blade moves into the inactive arc (dead band),
and the steam enters the blade passage, for an instant there is a force
produced on the pressure surface of the vane at exit from an active
arc (Fig. 5.10.8). There is no counterbalancing force on the suction
surface because that vane is now in the inactive arc, and has no
steam present. Therefore, there is a small unbalanced force +dF2
acting on the blade. This is termed the overshoot effect. In addi-
tion, for the next short interval of time, there is a small negative pres-
sure on the blade due to the flow through the passage. This effect is
shown as -dF3 in Figure 5.10.8.

Rotational stiffening effect


Large blades have natural frequencies that are lower than the
short blades. However, they accumulate a considerable number of
cycles, and will fail by, or be driven to, failure by some primary dam-
age site within relatively short periods of time. Also these blades are
tunedi.e., their manufactured frequency is adjusted to fall
within a band of acceptability before installation.

For these large blades, their centrifugal force has no effect on


their mass, but it has a considerable stiffening effect. Therefore, as
speed of rotation increases, the natural frequencies of the blades
increase. The Campbell equation for centrifugal stiffening is devel-
oped in the following manner.

Consider a particle of mass m, with an elastic support so stiff-


ness Rs is required to produce unit deflection. Its natural frequency
fo is expressed by:

1 Rs
fo =
2 m

541
Turbine Steam Path Maintenance and RepairVolume One

If this same particle is supported in another manner with an elas-


tic stiffness Rc, its new frequency fc will be given by:

1 Rc
fc =
2 m

If both these stiffening effects are applied simultaneously, the fre-


quency will then be fr, and can be found from:

1 Rs + Rc
fr =
2 m

Then if Rs is the stiffness furnished by elastic supports, and


Rc is the stiffness due to centrifugal effects, and assuming the lat-
ter is proportional to N, the speed of revolution is rps.

2 2
Rc = 4 mN

Using the equation for fo, and eliminating Rs gives:

2 2
fr = fo + N

In these expressions is an arbitrary constant, and the value of


usually lies between 2.0 and 3.0 for the lower modes.

It is clear from this expression that as the speed of rotation


increases, the natural frequency of the blade will increase. Therefore,
it is necessary to be able to represent this change of frequency in a
manner that enables the effect on any blade to be predicted under
any variation of speed. This is what the Campbell Diagram achieves.

The factor. Most last stage blades have a tip/root diameter


ratio that does not vary significantly from one manufacturer to anoth-
er. Also, most last stages are designed for a substantial 50% reaction
at their mean diameter, and a variation of reaction along the length

542
Steam Path Damage Induced by Material Degradation

of the blade to account for the radial flow or vortex effect of the
steam. This means the blade inlet and discharge angles will not be
significantly different from one manufacturers element to another.
There will also be minor section differences that reflect the research
each manufacturer contributes to development of specific designs.
Such differences are not expected to cause significant effects on the
value.
fo < 127 >
fo2 < 16129 >
N2 < 3600 >

1.8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4


N2 6,840 7,200 7,920 8,640 9,360 10,080 10,800 11,520 12,240
fr 150.4 152.7 155.1 157.4 159.6 161.9 164.1 166.1 168.4
Table 5.10.1 Operating Frequencies as a Function of Fundamental Frequencies fo,
N = 60 cps

A sensitivity analysis is made for a blade for different values of


from 2.0 to 3.0 in Table 5.10.1. The variation of blade frequency
at 60 Hz, for a blade, having a standstill frequency of 127 cps is
shown in Figure 5.10.9.

Design operating frequency


180

170
Blade frequency - cps

160

150

140

130

120
10 20 30 40 50 60 70
0
Operating frequency - rps

Figure 5.10.9TheFigure 5.10.9


rotational stiffening effect on
The rotational stiffening effect on a blade for ' ' values
a blade for values from
from 2.0 2.0 to 3.0.
to 3.0.

543
Turbine Steam Path Maintenance and RepairVolume One

Even if the blade section geometry were to be so distorted in


manufacture that its value was off design by 10%, it can be seen
from Figure 5.10.9 this would not alter the operating frequency too
much. In fact, a 10% variation represents more than twice the value
that could be expected. As an example consider a blade produced
with a 5% error along its entire length, having an effective mean
depth error of 5%.

The blade frequency can be found from:

Ie
fo = Ky .
Ae

where:

fo = Blade frequency in cps


Ie = The effective section modulus along length
Ae = The effective section area along length
Ky = A constant

The effective width and depth are shown in Figure 5.10.10.

We

Te

Figure 5.10.10The
Figure 5.10.10effective
width We and
The effective depth
width 'We'Te
andof the
depth
vane profile.
'Te' of the vane profile.

544
Steam Path Damage Induced by Material Degradation

3 3
We . Te
foerror = Ky .
We . Te

It can be seen that a 5% variation in the factor will have a


marginal effect on operating frequency fr. However, it is of value to
evaluate the effect of variation on a typical last stage blade, and
determine how at its extreme values, it could place a blade into a
risk category. As an example, consider a blade with a predicted
standstill frequency of 127 cps, for operation at 60 Hz, with a 2-pole
generator. The divergence of the centrifugal stiffening curve can be
shown on the Campbell Diagram, Figure 5.10.11. The values of fr
are shown calculated in Table 5.10.2.
'operating speed'.
Blade frequency at

Figure 5.10.11The variation of the value


Figure of the operating speed fundamental
5.10.11
frequency for a blade
The variation of thewith a standstill
value frequency
of the operating of fundamental
speed 127 cps, as afrequency
function of
. for a blade with a standstill frequency of 127 cps, as a function of ' '.

The vibration signature of two identical L-1 blades are shown


in Figures 5.10.12(a) and (b). From these two figures it can be seen
there is relatively close agreement at the lower valuesthe funda-
mental and first two harmonicsbut a degree of divergence at the
higher values. This is typical for normal longer blades, and the man-
ufacturer will under certain circumstances adjust these elements, if
there is a need for such tuning.

545
Next Page

Turbine Steam Path Maintenance and RepairVolume One

N 5 10 15 20 30 40 50 60 70
N2 25 100 225 400 900 1600 2500 3600 4900
< 2.0 >
N2 50 200 450 800 1800 3200 5000 7200 9800
fr 127.2 127.8 128.6 130.1 133.9 139.0 145.4 152.7 161.0
< 2.0 >
N2 75 300 675 1200 2700 4800 7500 10800 14700
fr 127.3 128.2 129.6 131.6 137.2 144.7 153.7 164.1 175.6
Table 5.10.2 Blade Frequency as a Function of Rotor Speed = 2.0 and 3.0

Figure 5.10.12The vibration signatures of two nominally identical blades


for an L-1 stage. The fundamental and lower harmonics are shown.

546
Chapter

6
Steam Path Damage
and Deterioration
from the Deposition
of Contaminants

INTRODUCTION
The thermal-power cycle is a complex arrangement of individual
pieces of equipment. Many act to control the expansion of steam;
others come into contact with other forms of water, which may con-
tain undesirable contaminants. These contact locations have the
potential to contaminate the working fluid. Other units operate on
steam removed directly from the earth; with no effort made to clean
the steam to the extent the contaminants transported by the geother-
mal steam are removed before entering the steam path.

Contaminants carried by the parent steam into the turbine can be


either chemically aggressive or non-aggressive. Aggressive compounds

607
Turbine Steam Path Maintenance and RepairVolume One

considerably influence the extent to which the steam path deteriorates


(in terms of its structural integrity). Both can influence the efficiency
with which the steam expands through the unit.

There are two major concerns with steam borne contaminants:

reduction in efficiency due to surface deterioration of the


steam path elements, and any resulting frictional losses that
occur

reduction in structural integrity resulting from corrosive


action on the various components of the unit

There also can be considerations of the possible change in rotat-


ing blade natural frequencies if the deposited matter becomes of sig-
nificant mass.

The extent to which surface frictional losses influence efficiency


deterioration is a function of the Reynolds Number associated with
the flow across the surface being considered. These losses are in turn
dependent upon the steams physical properties and flow velocities
at the stage locations. They are also influenced by the dimensional
characteristics of the row. Because the velocity of the steam flowing
through an expansion passage is changing (due to both stage reac-
tion and friction), the actual steam velocitywhich establishes the
Reynolds Numbervaries throughout the flow passage. This vari-
ability is overcome by using data derived by experimentationdata
normally developed by the manufacturers and applicable to their
families of profiles.

While the best data (in terms of the actual Reynolds Number) will
be determined by the manufacturer for operation with the same pro-
files at similar steam conditions, sufficient accuracy can be obtained
by calculating Reynolds Numbers based on discharge velocities,
which is the procedure normally used by the manufacturer.

608
Steam Path Damage and Deterioration from the Deposition of Contaminants

In the high-pressure and reheat sections of a unit, deposition pat-


terns are substantially more regular and deposits more evenly spread
than those occurring in the low-pressure section/latter-stage blades,
where a considerable amount of deposition occurs from the water
phase.

A number of deposition patterns can be observed in any blade


row, and it is remarkable that the patterns on any row are so similar
from blade to blade. This is an important consideration when attempt-
ing to determine a frictional loss for any row, because if there was
considerable variation from blade to blade it would be particularly
difficult, if not impossible, to define a mean effective surface rough-
nessa major characteristic for establishing losses that occur. In fact,
the judgment of mean surface roughness in a unit removed from serv-
ice is judgmental on the part of the engineer making the assessment,
as variation in depositions from inlet to discharge, from root to tip,
and on the pressure and suction faces must be considered.

The deposition of compounds that occur on the blades is of two


basic types:

compounds insoluble in water, which can only be removed


by mechanical means

soluble compounds that can be removed by immersion or


washing in water

The second characteristic can be advantageous under certain cir-


cumstances.

A more insidious consequence of chemical deposition on steam


path elements is that some of these compounds are chemically
aggressive and have the potential to come out of solution and
deposit on the internal surfaces of the unit as steam conditions
reduce on expanding through the various blade rows. Many of these
compounds have the potential to migrate and accumulate in hide-
outs, where under suitable environmental conditions they become

609
Turbine Steam Path Maintenance and RepairVolume One

chemically aggressive, causing reactions that eventually lead to


component deterioration and ultimately failure.

Steam conditions at any stage are a function of boiler delivery


conditions and the load the unit is generating. Therefore, as there are
electrical load changes, conditions on certain blade rows will mod-
ify, moving the steam conditions at any location into either the
superheated or saturated regions. Under such operating conditions
the deposition, collection, and concentration mechanisms in place
will provide the environment for various forms of corrosion.

This chapter will examine these two phenomena induced by


deposition, and consider the effects on efficiency degradation and
the potential this deposition provides to promote mechanical failure.

SOURCE OF
STEAM PATH IMPURITIES
Operating engineers need to address operating problems associ-
ated with chemical ingress and contamination on a continuing basis.
These compounds can gain access to the unit through various
sources and mechanisms. Some occur as a result of mechanical
damage within other components of the steam power cyclemost
notably the condenser. Other sources are water treatment plants,
where careful control of the treating process is required.

Oxygen can also gain access at many points in the cycle where
sub-atmospheric conditions exist. Units are normally equipped to
remove any air that enters the system before it can cause significant
corrosive damage, or support a chemical reaction. However, a cer-
tain amount can, and often does, gain access to the steam path.

610
Steam Path Damage and Deterioration from the Deposition of Contaminants

External sourcesCarried in at system leaks


A common source of contaminants in the steam path is infiltra-
tion by untreated (raw) water into the system. This commonly occurs
within the condenser, which is used to produce an operating vacu-
um. Cooling water may be taken from lakes, rivers, or the sea (in
coastal installations). There are also river installations that are tidal.

It is especially necessary to consider the effects of installations


where saline water (seawater) exists, because through access to the
condenser, saline can produce a significant ingress of contaminants.
Seawater contains high levels of many substances, most notably salt
(NaCl), which can be highly corrosive if it gains access to the steam
path.

System internalGenerated by water


treatment systems
Modern power generating systems require water steam in the
boiler to be demineralized and as pure as possible. This requirement
introduces the need for water treatmenta process that will cause the
introduction of excess chemicals into the unit steam path if not con-
trolled within close tolerances. Certain of these treatment substances
have the capability of forming chemically undesirable products.

Contaminated attemperating spray water


Water used to control the inlet temperature from the superheater
or reheater portion of the boiler must be taken from a reliable and
clean source (normally the feedwater).

Contaminated exhaust hood spray water


Water is used in the low-pressure hoods to control steam and
blade temperature at low loads when steam will reenter the last stage

611
Turbine Steam Path Maintenance and RepairVolume One

blades with the potential to cause trailing edge erosion (see chap-
ter 3). This spray water should be taken from a clean, reliable source.
Often water taken from the condenser hot well contains impurities
and dissolved oxygen that can initiate and also promote corrosion.
In certain nuclear units this water can be taken from an auxiliary
source.

Use of sodium sulfide as an oxygen scavenger


Sodium sulfide is often used as an oxygen scavenger in high-
pressure boilers. Decomposition of the sodium sulfide can produce
hydrogen sulfide H2S in the early moisture region of the unit.

Use of cleaning material at outages


Cleaning fluids may contain unacceptable levels of caustics,
chlorides, or sulfur when used for removing deposits and other
cleaning activities during a unit outage. This may be acceptable for
the cleaning process, but if components are not rinsed thoroughly of
cleaning fluid and such residues remain, they have the potential to
collect and produce corrosive compounds throughout the unit.

Manufacturing and repair lubricants


Steam turbine manufacturers employ sophisticated materials and
production processes, many requiring the use of cutting lubricants
and fluids containing compounds of sulfur, chlorine, phosphorus,
and magnesium. These substances, even in trace quantities, should
be removed from component surfaces before acceptance and before
they are assembled and/or shipped to site.

Fluids are required to both lubricate and cool the metal cutting
process. Such fluids can be either oil- or water-based compounds.
The most suitable are dependent upon the material being cut, the cut-

612
Steam Path Damage and Deterioration from the Deposition of Contaminants

ting rate, and the depths of cut employed. Very few components of
the steam turbine can be machined without the use of a cutting/cool-
ing fluid. Therefore, residuals of these compounds are always poten-
tially present on the surface of various components. This is a condi-
tion that can exist when the unit is new or when it is returned to serv-
ice after repair, but such impurities can also be introduced when
replacement parts are installed at a maintenance outage.

Shipping and storage preservatives


When components are shipped to site, many of them require
protection during transportation, including steps to prevent atmos-
pheric corrosion. If these components are to be stored at site during
installation, and possibly placed in long-term inventory, they will
require corrosion protection.

Materials used for such protection must be suited to the antici-


pated life, and may be of a nature that if not cleaned completely
before installation, could introduce chemically undesirable com-
pounds into the system or even attack the component during storage.
Figure 6.2.1 shows blades removed from inventory for installation
that have suffered corrosive attack.

Dissolved from system equipment


Another source of contaminants carried into the steam path are
those produced within other pieces of equipment comprising the
power cycle. Most notable of these are iron and copper oxides pro-
duced as byproducts of dissolving metals from tubes and equipment
surfaces. This includes boiler tubing and regenerative feedwater
heaters, where high-temperature, high-pressure water comes into
contact on a continuing basis as the unit produces power.

613
Turbine Steam Path Maintenance and RepairVolume One

Figure 6.2.1Corrosive damage formed on new blade elements


while they were in storage.

Geothermal
The steam used in units supplied for geothermal application tends
to be present at pressures relatively lownormally lower than the
inlet conditions used for the nuclear cycle. This steam is also con-
taminated by minerals and compounds dissolved in it from the earth.

Table 6.2.1 shows the mix of gases, water, and impurities found
at different sites. However, the contaminants can vary considerably
from site to site, both in constituents and concentration. For direct-
cycle units, these contaminants are carried into the steam path,
where many are deposited on the internal surfaces, causing levels of
surface roughness and corrosion considerably greater than normally
experienced in fossil or nuclear cycles.

614
Steam Path Damage and Deterioration from the Deposition of Contaminants
Table 6.2.1Geothermal Steam-Water, Gas and Impurity Content
615
Turbine Steam Path Maintenance and RepairVolume One

In terms of deposition of impurities, geothermal units provide an


entirely discrete set of considerations. This is because there can be
deposits in these units completely unlike those found in units used
for cycles of the more conventional fossil and nuclear fueled designs
(Fig. 6.2.2).

Figure 6.2.2The deposits on a geothermal stationary blade row. These deposits, while
not corrosive, had a very significant effect on strage efficiency. The effect of a stay bar
holding one stationary vane in place, and the flow divertion, and therefore deposition
pattern can also be seen.

Geothermal units are arranged to be either direct cycle or indi-


rect cycle. In direct-cycle units, naturally occurring steam removed
from the earth is passed through the steam path, often with separa-
tors interposed between the steam supply and the turbine. The sepa-
rators remove suspended moisture and solid particulate from the
steam. There is little or no effort made to remove suspended and dis-
solved chemical impurities.

Indirect-geothermal cycles employ a heat exchanger vessel


between the geothermal steam and the working fluid. In these units,

616
Steam Path Damage and Deterioration from the Deposition of Contaminants

the working fluid should have a purity comparable to that of fossil and
nuclear units. However, these units can be subject to the intrusion of
contaminants or impurities as the result of tube failures in the heat
exchanger vessels that dependent upon the corrodents carried in, can
be relatively common. In indirect cycles, the heat exchange vessels
can themselves suffer heavy deposition and corrosive damage.

Geothermal steam tends to be saturated, containing at best only


a few degrees of superheat. Therefore, steam path components are
almost all in the saturated region, with the probability of continuous
washing. Water present on steam path surfaces may itself contain
impurities that are deposited and detrimental to the total perform-
ance of the unit. However, it can have some beneficial effect in terms
of efficiency, as it will tend to remove some level of the deposited
compounds.

Water quality
After initial unit start up and commissioning, the majority of the
contaminants found in the steam path are those carried in by the
steam. Therefore, control of steam quality is essential.

Note: During initial start up and commissioning, any contami-


nants due to manufacture, storage preservatives, and other once-
only circumstances are removed almost entirely. However, when a
component has been removed for rebuild or repair, it is a good idea
to examine and possibly clean it before its return to service.

Control of steam quality is obviously synonymous with control of


boiler water/steam. The recommendations outlined in Table 6.2.2
represent the upper limits considered necessary to provide steam of
sufficient purity that cleaning at maintenance outages should pre-
clude the incidence of major deposits, scaling or corrosive action.
Operation beyond these upper limits should be avoided but if
exceeded, immediate corrective action should be implemented.

617
Turbine Steam Path Maintenance and RepairVolume One

Contaminant Normal Limiting Condition


Operation 2 Weeks 24 Hours
Dissolved Oxygen < 10 10 - 30 30 - 100
Sodium < 5 5 - 10 10 - 20
Chlorides < 5 5 - 10 10 - 20
Silica < 10 10 - 20 20 - 50
Copper < 2
Iron < 20
Sodium Phosphate 2.3 - 2.7
Sulfides and Sulfates: Less than detectable, should be analyzed at least
once a week.

Table 6.2.2Steam Purity Recommendations, in parts per billion (ppb)

Figure 6.2.3The result of a large sea water ingress at the condenser, and deposition
of salt (NaCl) on a rotating blade row.

618
Steam Path Damage and Deterioration from the Deposition of Contaminants

Unfortunately, it is possible that a unit may operate for years


below the recommended levels, and then a single system disruption
of several hours duration can cause a contaminant ingress sufficient
to produce a deposition rate highly detrimental to the continued reli-
able operation of the unit.

Figure 6.2.3 shows a unit using seawater for cooling in the con-
denser. This unit operated reliably until a severe leak of several hours
duration occurred. The consequent deposition of contaminants war-
ranted the unit be shut down and completely cleaned (including the
removal of several rows of rotating blades to allow cleaning of the
blade roots and the rotor portion of the root fastening). This was an
expensive outage, but not nearly as expensive (in terms of time and
material) as the possible need to reopen the unit within a short peri-
od of return to service and correct blade damage (and possibly rotor
corrosion), with the consequent purchase of replacement blades.

THE COMPOSITION
OF DEPOSITS
The composition and concentration of the compounds deposited
on the steam path component surfaces are dependent upon their ori-
gin. Irrespective of the level of care taken by plant operators to limit
their presence, impurities will gain access from various sources and
be present in the steam. There will be a persistent deposition, possi-
bly slow, throughout the operating life of the unit. To best categorize
these life-long compounds, consider those that gain access through
leakage into the cycle, from some secondary process of heat
exchange or system maintenance, and those present as a conse-
quence of some form of treatment undertaken on the working fluid.

619
Turbine Steam Path Maintenance and RepairVolume One

Probably the most frequently found corrosive agents affecting the


blade pathand potentially the most damagingare sodium chlo-
ride and sodium hydroxide. The introduction of caustic contami-
nants into the water/steam cycle is generally attributed to the ingress
of common salt (NaCl) into the system, and the dissociation of this
compound into sodium (Na) and chlorine (Cl) ions, which can then
recombine with ions of hydroxide (OH) and hydrogen (H) to form
caustic sodium hydroxide (NaOH) and hydrochloric acid (HCl).

Since these compounds are normally controlled at extremely


low levels, it is necessary to consider the mechanisms by which their
deposition and concentration is increased to a level initiating some
form of corrosion damage.

Once it gains access, sodium chloride readily dissolves in water.


Sodium hydroxide, once formed, can be transported by the steam in
the boiler feedwater through the feed heating train to the boiler. In
the superheater, residual water is evaporated, leaving molecules of
sodium hydroxide in equilibrium with the steam molecules. These
sodium hydroxide molecules in their equilibrium condition then
enter the steam turbine and flow through the steam path with the
expanding steam. As temperature and pressure reduce, and the
steam passes into the saturated region, the concentration will fall,
i.e., the molecules of sodium hydroxide will capture water mole-
cules to dilute the concentration and maintain an equilibrium con-
dition.

Note: To get into the water/steam system the sodium chloride is


more likely to be dissolved in waterit will be about 90% ionized.
In the superheated steam the residual sodium chloride and the sodi-
um hydroxide will exist as ionized pairs (Na+Cl- and Na+OH-) in a
weak association, and will be precipitated onto the surfaces of the
blade path. As this action takes a finite time, some ion pairs will be
carried down into the wet region and be absorbed into the water. If
not deposited on the low-pressure section surfaces, the absorbed

620
Steam Path Damage and Deterioration from the Deposition of Contaminants

sodium chloride and sodium hydroxide will be transported into the


condenser hot well.

The melting point of sodium hydroxide is a little above 600F. At


temperatures below this, these ions will come out of solution and
deposit on the steam path elements, as the water droplets containing
them come into contact with the steam path components. These sodi-
um hydroxide molecules, once deposited with the water, can flow or
wash into, collect, and be retained within various crevices or hide-
outs formed in the steam path. These collection regions are often not
easily accessible for cleaning, or removal by the washing action of
the water phase of the working fluid. There is a tendency for these
deposits to collect, concentrate and remain in these hideouts.

When pressure and temperature within the steam path change,


the deposited molecules will attempt to reach equilibrium condi-
tions. This may not always be readily attainable, and may take
extended periods of operation at the new conditions to achieve it.
This is because of a time delay before the metal temperatures of the
major components can achieve a new operating condition.
Depending on their location within the steam path, the solution of
sodium hydroxide in these hideouts will take varying amounts of
time to achieve a new equilibrium condition. These time delays may
leave a solution of sodium hydroxide in a particularly active state for
extended periods each time a pressure and temperature change
occurs within the steam path.

This increased activity is also possible when a unit is operating at


off-design steam conditions, and changes in inlet steam temperature
cause a fluctuation or variation in stage temperatures. Under these
conditions, components and regions of the unit that should be in a
no sodium hydroxide region are moved into an active sodium
hydroxide region and can, as a consequence, suffer corrosive attack.

In areas of high temperature and pressure (superheated), stress


corrosion due to sodium hydroxide activity cannot occur too readily.

621
Turbine Steam Path Maintenance and RepairVolume One

This is due principally to the fact that steam and component temper-
atures are above the melting point of sodium hydroxide. The higher
density of the steam in these regions, and its ability to prevent sodi-
um hydroxide molecules from depositing in quantities sufficient to
cause any corrosive action, also plays a role.

The number of failures and instances of corrosive damage found


in the low-pressure regions of the steam turbine (due to sodium
hydroxide) indicates components in this region are vulnerable, if
corrosive agents can adhere to surfaces and find hideouts into which
they can migrate, collect, and concentrate.

In assessing possible changes in concentration of corrodents


influenced by steam conditions at various locations, the following
factors should be considered:

At full load operation, low-pressure stages vulnerable to cor-


rosive attack may be operating in the saturated region

As load is reduced, there is a greater degree of throttling on the


control stage or at the control valves on throttle-controlled units. This
action modifies the expansion line, moving it to the right on the
Mollier Diagram, with the susceptible stages moving from the satu-
rated to the superheated region. The expansion in these stages now
occurs in the superheat region, and some of the rows are exposed to
superheated steam.

This condition change is shown on the Mollier Diagram (Fig.


6.3.1). In this unit, the expansion lines are shown for full load con-
dition A,BE, and at part load A, BmEm. The effect of partial
arc admission or throttling is at the portion of the unit moving the
high-pressure section exhaust from B to Bm. The final exhaust
from the low-pressure section is maintained at Pxd, but the dryness
fraction has changed from E to Em, which is substantially drier.

The stages in the low-pressure section where the transition from


superheated to saturated conditions are shown as stage L-N-K. The

622
Steam Path Damage and Deterioration from the Deposition of Contaminants

effect of throttling is to move these stages to conditions Lm-Nm-


Km. This will modify the dryness fraction from Xid to Xdp at full
load and to superheated to Xde at part load. The implication of this
condition change is that throttling, and even the degree of superheat
on a stage can vary as the load changes. There is the potential for
concentration of any corrodent, and while operating in the part-load
condition, corrosive action can initiate and continue.

Failure to maintain steam temperatures from the boiler super-


heater and/or reheater at design conditions

Pr
C Cm

Pi
A
Ti Km
K
Pr

B Bm
Dm

LP Inlet

Sat. X = 1.0
Pcu
Pcl

Xid Pip Wilson


L Line Tsup
Pid
sup

Xdp
Pde Pxd
K Km
N Xde Em
Nm
Pdp E

Figure 6.3.1Showing Figure 6.3.1


the effect of expansion stage condi-
tions when initial
Showing steam
the effect conditionsstage
of expansion change.
conditions when
initial steam conditions change.

623
Turbine Steam Path Maintenance and RepairVolume One

The amount of initial and reheat temperature reduction sensed


by any unit depends upon the type of boiler and the load at which
the unit is operating.

In general, lowering the boiler delivery temperatures while main-


taining pressure at or close to design values lowers steam conditions
throughout the low-pressure section. This temperature reduction can
possibly lower early stages of this section into the wet or active sodi-
um hydroxide region, at which condition any sodium hydroxide
residue can become chemically active and cause corrosion.

Consider the unit expansion shown in Figure 6.3.2. Steam is


admitted to the unit at condition A and expands through the unit
with steam conditions represented by line A...D. With the reheat
temperature at the same level as the initial, the final dryness fraction
will be Xd. If, however, the reheat steam temperature is lowered by
an amount T, then the final dryness fraction will be Xdm and
stages designed for the superheat section of the expansion will be
operating in the wet region.

At light load (below 25%) the steam flow is principally


through the outer flow portions of the steam path

This is particularly true in the longer, low-pressure blades. This


effect causes re-circulation through the lower portions of the rotating
blade row, causing windage and frictional heating of the latter stage
blades at the root and lower sections of the vane. This effect is shown
in Chapter 3, Figure 3.9.7. If water sprays are not operative, or the
temperature of the steam and metal has not increased to a significant
degree, there will be a chemically active corrodent near the root and
lower vane portions of the blades.

At flow sealing constrictionsdiaphragms, shaft ends, and